Label-free sequencing approaches and applications for heterogeneous samples

Abstract

The present disclosure provides methods and chips for detection and analysis of biological samples (e.g., analytes) in a label-free manner. Chips of the present disclosure may comprise a resonator. The resonator may comprise a nanostructure configured to enhance resonance for vibrational spectroscopy, and to attach a biological molecule of a biological sample to said functionalized nanostructure.

Description

LABEL-FREE SEQUENCING APPROACHES AND APPLICATIONS FOR HETEROGENEOUS SAMPLESSEQUENCE LISTING

[0001] This application contains a Sequence Listing that has been submitted in electronically in Extensible Markup Language (.xml) format and is hereby incorporated by reference in its entirety. Tire XML copy, created on January 27, 2025, is named 126041 834173.xml, and is 17,316 bytes in size.BACKGROUND

[0002] Sample analysis methods and devices enable deep analysis of materials, especially biological materials. Analysis of non-labeled samples can provide insights into the properties of the sample and can improve acquisition speed and sample fidelity.SUMMARY

[0003] The disclosure provides methods for processing heterogeneous mixtures of molecules. The heterogeneous mixtures of molecules may be from biological samples. Biological samples may include proteins, nucleic acid, cells, or other components described herein. The heterogeneous mixtures of molecules may be from chemical samples. The method may include chemical means to selectively bind one or more molecule. Hie methods may include bringing one or more molecule to a surface (e.g., chip). The one or more molecule brought to the substrate may be the selectively bound one or more molecule. The method may include attaching one or more molecule to a substrate (e.g., chip). The one or more molecule attached to the substrate may be the selectively bound one or more molecule. Tire one or more molecule may be attached to the surface in a specific orientation. The attachment orientation of the one or more molecule may be chemically based.

[0004] The disclosure provides substrates (e.g., chips) for use in providing the composition of one or more molecule. The methods disclosed herein may use substrates (e.g., chips) to provide a readout of the composition of one or more molecule.

[0005] Aspects disclosed herein provide methods of identifying a biological molecule, comprising: (a) providing a resonator comprising a nanostructure; (b) bringing said resonator in contact with said biological molecule under conditions sufficient to pennit said biological molecule to couple to said nanostructure; (c) exposing said biological molecule to a first light; (d) subsequent to (c), measuring a second light emanating from said biological molecule; and (e) using at least said second light to identifysaid biological molecule. In some embodiments, said biological molecule is a single biological molecule. In some embodiments, said resonator comprises a first element and a second element separated by a gap, wherein said nanostructure is disposed in said gap. In some embodiments, said gap is a nanogap. In some embodiments, said resonator is part of a chip, and wherein said method further comprises providing a mixture comprising said biological molecule to said chip. In some embodiments, said nanostructure is functionalized. In some embodiments, said nanostructure is chemically functionalized. In some embodiments, said functionalized nanostructure is functionalized with one or more moieties that reduce non-specific binding. In some embodiments, said functionalized nanostructure is functionalized with one or more of the following moieties: (i) an alkyl-thiol; (ii) an alkyl -oxysilane; (iii) an alkyl-selenol; (iv) an oxygen containing alkyl-thiol; (v) oxygen containing alkyl-oxysilanes; (vi) oxygen containing alkyl- selenol; (vii) zwitterionic containing thiols; (viii) zwitterionic containing oxysilanes; (ix) zwitterionic terminated selenol; (x) cyclic containing thiols; (xi) cyclic containing oxysilanes; (xii) cyclic containing selenol; (xiii) azide containing thiols; (xiv) azide containing oxysilanes; (xv) azide containing selenol; (xvi) alkyne containing thiols (xvii) alkyne containing oxysilanes; (xviii) alkyne containing selenol; or (xix) molecules containing all or some of the following: (a) oxygen groups; (b) alkyl groups; or (c) a zwitterionic group. In some embodiments, said chemical functionalization comprises reducing a density of a tethering molecule in a non-specific binding reducing agent. In some embodiments, said chemical functionalization comprises backfilling into an established matrix. In some embodiments, said chemical functionalization comprises co-depositing with an established matrix. In some embodiments, said biological molecule comprises a protein or peptide. In some embodiments, said protein is a major histocompatibility complex (MI-IC) protein. In some embodiments, said MHC protein comprises a class I peptide. In some embodiments, said MHC protein comprises a class II peptide. In some embodiments, said protein comprises a protein isoform. In some embodiments, said protein comprises an antibody. In some embodiments, said biological molecule comprises a biologic. In some embodiments, said biological molecule comprises a nucleic acid molecule. In some embodiments, said nucleic acid molecule comprises a deoxyribonucleic acid (DNA) molecule. In some embodiments, said nucleic acid molecule comprises a ribonucleic acid (RNA) molecule. In some embodiments, said biological molecule comprises a sugar molecule. In some embodiments, said biological molecule is configured to be detected, and wherein a signal intensity associated with said biological molecule provided on said chip is at least two times higher than a signal intensity associated with a biological molecule not provided on said chip. In some embodiments, said biological molecule is identified by cleaving an amino acid from said protein or peptide. In some embodiments, said amino acid is cleaved from tire N-terminus of said protein or peptide. In some embodiments, cleaving said amino acid further comprises serial introduction of chemicals capable of performing said cleaving. In some embodiments, the method further comprises generating asynthetic spectrum. In some embodiments, the synthetic spectrum is a close approximation to an empirical spectrum. In some embodiments, the synthetic spectrum is developed by signal processing pipelines and machine learning models. In some embodiments, the synthetic spectrum is used to augment datasets, wherein the augmented data sets further comprises training a machine learning model. In some embodiments, machine learning models are trained to identify said biological molecule based on a spectrum. In some embodiments, the machine learning models are trained with any one of amino acid sequences, nucleic acid sequences, post translational modifications, secondary7structures, higher-order structures, or any other chemical composition or property attributes. In some embodiments, a spectrum is generated through chemical, electrical, or physical perturbations. In some embodiments, perturbations are generated from heating, charge or electric field perturbations, pH perturbations, or chemical binding and cleavages. In some embodiments, the method further comprises estimating a three-dimensional structure of the biological molecule. In some embodiments, a machine learning model estimates the three- dimensional structure of the biological molecule. In some embodiments, the machine learning model estimates the three-dimensional structure based on post translational modifications. In some embodiments, the machine learning model estimates the three-dimensional structure based on chemical bond energies. In some embodiments, the post translational modification and / or the chemical bond energies generate at least 10% higher specificity and / or confidence values than existing structure prediction algorithms (e.g., AlphaFold, AlphaFold2. PepFold). In some embodiments, the machine learning model predicts a density or a number of molecules coupled to said nanostructure. In some embodiments, predicting comprises generating probabilities for the number of molecules coupled to said nanostructure. In some embodiments, said density or said number of molecules corresponds to a concentration of said molecules. In some embodiments, the machine learning model predicts type or types of molecules bound to the nanostructure. In some embodiments, the method further comprises predicting the proportion of the types of molecules bound to the nanostructure. In some embodiments, the machine learning model generates a quantification estimate for each type of molecule analyzed over an entire array. In some embodiments, the method further comprises generating an optical readout. In some embodiments, the optical readout is based on a vibrational fingerprint of the biological molecule. In some embodiments, the optical readout is based on a signal intensity of the biological molecule. In some embodiments, the resonator is configured to enhance the vibrational fingerprint and / or signal intensity of the biological molecule. In some embodiments, the resonator enhances the vibrational fingerprint and / or signal intensity of a biological molecule coupled to the resonator at least two times as compared to a biological molecule not interacting with the resonator. In some embodiments, the resonator enhances the vibrational fingerprint and / or signal intensity of a biological molecule interacting with the resonator at least 102 times as compared to a biological molecule not interacting with the resonator. In someembodiments, the resonator enhances the vibrational fingerprint and / or signal intensity of a biological molecule interacting with the resonator at least 103times as compared to a biological molecule not interacting with the resonator. In some embodiments, the resonator enhances the vibrational fingerprint and / or signal intensity of a biological molecule interacting with the resonator at least 105 times as compared to a biological molecule not interacting with the resonator. In some embodiments, the nanostructure is metallic. In some embodiments, the resonator is an optical resonator. In some embodiments, the resonator is a dielectric resonator. In some embodiments, the resonator is an optical dielectric resonator. In some embodiments, the resonator is a semi conducting resonator. In some embodiments, the second light is a different frequency than the first light. In some embodiments, said biological molecule comprises no more than ten biological molecules. In some embodiments, said protein comprises a human leukocyte antigen (HLA). In some embodiments, said HLA a class I peptide. In some embodiments, said HLA a class II peptide. In some embodiments, the identifying of a molecule is performed de novo. In some embodiments, the synthetic spectrum is developed by using ab-initio calculations. In some embodiments, the method further comprises generating the synthetic spectrum using ab-initio calculations. In some embodiments, the synthetic spectrum is developed by using molecular dynamics simulations. In some embodiments, the method further comprises generating the synthetic spectrum using molecular dynamics simulations. In some embodiments, the resonator enhances the vibrational fingerprint and / or signal intensity of a biological molecule interacting with the resonator at least 1010 times as compared to a biological molecule not interacting with the resonator. In some embodiments, the resonator enhances the vibrational fingerprint and / or signal intensity of a biological molecule interacting with the resonator at least 1015times as compared to a biological molecule not interacting with the resonator. In some embodiments, the resonator enhances the vibrational fingerprint and / or signal intensity of a biological molecule interacting with the resonator at least 1020times as compared to a biological molecule not interacting with the resonator. In some embodiments, the nanostructure comprises a metal. In some embodiments, the metal comprises gold, platinum, palladium, copper, aluminum, titanium, chromium, or other noble metals, coinage metals, and any combination thereof. In some embodiments, said resonator comprises a first element and a second element separated by a gap, wherein said nanostructure is disposed atop said gap. In some embodiments, single molecule vibrational spectroscopy is for fingerprinting and / or sequencing of: (a) major histocompatibility complex (MHC) class I and / or II peptides; (b) antibodies and / or antibody fragments; (c) biologies, wherein said biologies include post-translational modifications and / or secondary / tertiary structure; (d) protein isoforms and / or protein fragments; (e) DNA and / or complementary DNA (cDNA); or (f) any combination thereof. In some embodiments, signal processing pipelines and machine learning models arc developed to synthesize spectra that closely approximate empirical spectra. In some embodiments, machine learningmodels are trained and used for spectra-to-molecule identification. In some embodiments, each input data point is composed of a set or series of spectra generated through chemical, electrical, or physical perturbations. In some embodiments, a 3 -dimensional structure of the biological molecule is estimated using a model with input comprising PTMs and chemical bond energies. In some embodiments, a trained model would predict a density or number of molecules coupled to the resonator. In some embodiments, a trained model predicts a mixture or types of biological molecules bound to the chip. In some embodiments, machine learning models are trained to provide quantification estimates for each analyte type present in an assayed sample over an entire array. In some embodiments, said nanostructure disposed atop said gap is disposed on a top face of a resonator element.

[0006] Aspects disclosed herein provide chips for identifying a biological sample, said chip comprising a resonator comprising a nanostructure configured to attach a biological molecule of said biological sample to said nanostructure. In some embodiments, said biological molecule is a single biological molecule. In some embodiments, said resonator comprises a first element and a second element separated by a gap, wherein said nanostructure is disposed in said gap. In some embodiments, said gap is a nanogap. In some embodiments, said resonator is part of a chip, and wherein said method further comprises providing a mixture comprising said biological molecule to said chip. In some embodiments, said nanostructure is a functionalized nanostructure. In some embodiments, said functionalized nanostructure is chemically functionalized. In some embodiments, said functionalized nanostructure is functionalized with one or more moieties that reduce non-specific binding. In some embodiments, said functionalized nanostructure is functionalized with one or more of the following moieties: (i) an alkyl-thiol; (ii) an alkyl-oxysilane; (iii) an alkyl-selenok (iv) an oxygen containing alkyl-thiol; (v) oxygen containing alkyl-oxysilanes; (vi) oxygen containing alkyl -selenol: (vii) zwitterionic containing thiols; (viii) zwitterionic containing oxysilanes; (ix) zwitterionic terminated selenol; (x) cyclic containing thiols; (xi) cyclic containing oxysilanes; (xii) cyclic containing selenol; (xiii) azide containing thiols; (xiv) azide containing oxysilanes; (xv) azide containing selenol; (xvi) alkyne containing thiols (xvii) alkyne containing oxysilanes; (xviii) alkyne containing selenol: or (xix) molecules containing all or some of the following: (a) oxygen groups; (b) alkyl groups; or (c) a zwitterionic group. In some embodiments, said chemical functionalization comprises reducing a density of a tethering molecule in a non-specific binding reducing agent. In some embodiments, said chemical functionalization comprises backfilling into an established matrix. In some embodiments, said chemical functionalization comprises co-depositing with an established matrix. In some embodiments, said biological molecule comprises a protein or peptide. In some embodiments, protein is a major histocompatibility complex (MHC) protein. In some embodiments, said MHC protein comprises a class 1 peptide. In some embodiments, said MHC protein comprises a class 11 peptide. Insome embodiments, said protein comprises a protein isoform. In some embodiments, said protein comprises an antibody. In some embodiments, said single biological molecule comprises a biologic. In some embodiments, said biological molecule comprises a nucleic acid molecule. In some embodiments, said nucleic acid molecule comprises a deoxyribonucleic acid (DNA) molecule. In some embodiments, said nucleic acid molecule comprises a ribonucleic acid (RNA) molecule. In some embodiments, said biological molecule comprises a sugar molecule. In some embodiments, said biological molecule is configured to be detected, and wherein a signal intensity associated with said single biological molecule provided on said chip is at least two times higher than a signal intensity associated with a biological molecule not provided on said chip. In some embodiments, said protein comprises a human leukocyte antigen (HLA). In some embodiments, said HLA a class I peptide. In some embodiments, said HLA a class II peptide. In some embodiments, said biological molecule comprises no more than ten biological molecules. In some embodiments, single molecule vibrational spectroscopy is for fingerprinting and / or sequencing of: (a) major histocompatibility complex (MHC) class I and / or II peptides: (b) antibodies and / or antibody fragments; (c) biologies, wherein said biologies include post-translational modifications and / or sccondary / tcrtiary structure; (d) protein isoforms and / or protein fragments; (e) DNA and / or complementary DNA (cDNA); or (f) any combination thereof. In some embodiments, signal processing pipelines and machine learning models are developed to synthesize spectra that closely approximate empirical spectra. In some embodiments, machine learning models are trained and used for spectra-to- molecule identification. In some embodiments, each input data point is composed of a set or series of spectra generated through chemical, electrical, or physical perturbations. In some embodiments, a 3- dimensional structure of the biological molecule is estimated using a model with input comprising PTMs and chemical bond energies. In some embodiments, a trained model would predict a density or number of molecules coupled to the resonator. In some embodiments, a trained model predicts a mixture or types of biological molecules bound to the chip. In some embodiments, machine learning models are trained to provide quantification estimates for each analyte type present in an assayed sample over an entire array. In some embodiments, the nanostructure comprises a metal. In some embodiments, said resonator comprises a first element and a second element separated by a gap, wherein said nanostructure is disposed atop said gap. In some embodiments, said nanostructure disposed atop said gap is disposed on a top face of a resonator element.

[0007] Aspects disclosed herein provide chips having multi-resonant features and / or nanostructures. The chip may comprise unit cells, which are patterned across the chip in an array. A chip (e.g., photonic chip) may comprise unit cell designs as described herein (e.g., shape and size). Hie unit cell may comprise one or more photonic pillars. The photonic pillar of a unit cell may be a primary photonic pillar (e.g.. havingenhanced resonance and / or analyte attachment). A primary photonic pillar may comprise one or more nanostructures (e.g., antennae) as described herein (e.g., a single or multiple-antennae design of varying shapes and sizes). The antennae may be a metal, as described herein. The photonic pillar of a unit cell may be a secondary photonic pillar (e.g., not having enhanced resonance). Photonic pillars may comprise dielectric materials (e.g., silicon), dielectric spacers, dielectric fill layers, oxides, passivation layers, and / or glass wafers. A dielectric spacer layer may be an oxide, optionally silicon dioxide. A silicon dielectric layer may comprise crystalline silicon, amorphous silicon, or silicon nitride. In certain embodiments, a multi-resonant chip features a mirror or mirror-like layer for further spectral enhancement.

[0008] The design of multi -re sonant chips (e.g., number and ratio of photonic pillar types and antennae; size and shape of photonic pillars and antennae, unit cell dimensions, and materials can be optimized based on the excitation (e.g., a laser) and emission (e.g., Raman spectra) for the biological sample being investigated. Tire chip design (e.g., chip layering) may be as shown in any one of FIGS. 52A-58B, FIGS. 68A-68D, or FIGS. 69A, 69D, 69E, or 69F. Also provides are methods of manufacturing said chips. The manufacturing steps may comprise one or more of the steps as described herein, and as shown in FIG. 70. Such multi-resonant chips chips can be used in any method or use as described herein.

[0009] Aspects disclosed herein provide a method of determining a cellular response of a cell to a stimulus, comprising (a) providing a cell and a photonic chip, wherein the chip comprises a resonator; (b) attaching the cell to the photonic chip at or adjacent to the resonator; (c) exposing the cell to a first light; (d) measuring a second light emanating from the cell; and (e) using at least the second light to identify the cellular response of the cell. Machine learning models may be trained to deconvulate data from the method for determination.

[0010] Another aspect provides a Raman macromolecule, comprising (a) a Raman reporter molecule and (b) a barcoded peptide-MHC protein complex, optionally wherein the barcode is a DNA barcode. Another aspect provides a Raman macromolecule, comprising (a) a particle; (b) a Raman reporter molecule; and (c) a barcoded peptide-MHC protein complex, optionally wherein the barcode is a DNA barcode. Also provided are methods of making the Raman marcromolecules. Said Raman macromolecules may be for use in vibrational spectroscopic determination of (a) a cellular response of a cell to a stimulus; (b) a cellular state of a cell; (c) a cellular type of a cell; or (d) a cellular binding activity of a cell. Machine learning models may be trained to deconvulate data from the method for determination.

[0011] Another aspect provides a Raman macromolecule, comprising a peptide-pulsed antigen presenting cell (APC), comprising (a) an APC; (b); at least one antigen, wherein each antigen is an immunogenic peptide; and (c) a Raman reporter molecule, wherein the Raman reporter is attached to the APC or to the antigen. The APC may be pulsed with tire at least one antigen. Also provided are methods of making the peptide-pulsed APC. Said Raman macromolecules may be for use in vibrational spectroscopic determination of (a) a cellular response of an APC after stimulation with an antigen; (b) a cellular state of an APC after stimulation with an antigen; (c) a cellular type of an APC after stimulation with an antigen; or (d) a cellular binding activity of APC after stimulation with an antigen. Hie stimulus may be applied to the cell before attachment to the photonic chip, or after attachment to the photonic chip. Machine learning models may be trained to deconvulate data from the method for determination.

[0012] Aspects disclosed herein provide a method of sequencing activated and exhausted CD8+ T cells, the method comprising contacting CD8+ T cells with a Raman macromolecule as described herein.

[0013] Aspects disclosed herein provide means for pre-processing analytes (e.g., biological samples) for vibrational spectroscopy. Hie pre-processing may comprise one or more of: chemical modification of the analyte: fractioning the sample for vibrational spectroscopy; spotting or printing the sample for vibrational spectroscopy; attaching the sample for vibrational spectroscopy to a photonic chip; obtaining a vibrational spectra for the photonic chip (e.g., Raman spectroscopy); or analyzing the vibrational spectra Data, optionally wherein the molecular species and of the sample and quantity of the sample is determined.

[0014] Additional aspects and advantages of the present disclosure will become readily apparent to those skilled in this art from the following detailed description, wherein only illustrative embodiments of the present disclosure are shown and described. As will be realized, the present disclosure is capable of other and different embodiments, and its several details are capable of modifications in various obvious respects, all without departing from the disclosure. Accordingly, the drawings and description arc to be regarded as illustrative in nature, and not as restrictive.

[0015] All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference. To the extent publications and patents or patent applications incorporated by reference contradict the disclosure contained in the specification, the specification is intended to supersede and / or take precedence over any such contradictory material.BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application with color drawing(s) will be provided by the Office by request and payment of the necessary fee.

[0017] The novel features of the invention are set forth with particularity in the appended claims. A better understanding of the features and advantages of the present invention will be obtained by reference to the following detailed description that sets forth illustrative embodiments, in which the principles of the invention are utilized, and the accompanying drawings (also “Figure” and “FIG.” herein), of which:

[0018] FIG. 1 depicts an example of a processing workflow, according to some embodiments.

[0019] FIG. 2A depicts exemplary designs of an array of non-uniform features, according to some embodiments.

[0020] FIG. 2B depicts an example view of an array design, according to some embodiments.

[0021] FIGS. 3A-3F depict alternative designs for arrays of non-uniform features, according to some embodiments.

[0022] FIG. 4 depicts an example of a two-dimensional array of non-uniform features, according to some embodiments.

[0023] FIGS. 5A-5B depict examples of two-dimensional arrays, according to some embodiments.

[0024] FIG. 6A depicts tire portions of an array, according to some embodiments.

[0025] FIG. 6B depict an example of the field enhancement of an array not comprising a nanogap, according to some embodiments.

[0026] FIGS. 7A-7B depict an example of a field enhancement calculation and an inset zoom of an array comprising a plurality of gaps, according to some embodiments.

[0027] FIG. 8A depicts an example of an array with a single nanogap, according to some embodiments.

[0028] FIG. 8B depicts an example field profile for tire array of FIG. 8A, according to some embodiments.

[0029] FIG. 9 depicts an example of a chip comprising a detection region and a separation region, according to some embodiments.

[0030] FIGS. 10A-10C depict a pathway for analysis of the spectra of the present disclosure, according to some embodiments.

[0031] FIG. 11 depicts an example of tissue mapping with an array, according to some embodiments.

[0032] FIG. 12 depicts an example micrograph of a plurality of arrays, according to some embodiments.

[0033] FIG. 13 depicts a micrograph of an example array, according to some embodiments.

[0034] FIGS. 14A-14B depict micrographs of example pluralities of array, each array comprising a plurality of gaps, according to some embodiments.

[0035] FIG. 15 depicts an example of spectral measurement of an interaction, according to some embodiments.

[0036] FIG. 16 depicts sample Raman spectra, according to some embodiments.

[0037] FIG. 17 depicts a computer system that is programmed or otherwise configured to implement methods provided herein.

[0038] FIGS. 18A-18D show additional examples of fabricated arrays, according to some embodiments. FIGS. 18A and 18B depict example fabricated structures with gaps along the full resonator. FIGS. 18C and 18D depict example fabricated structures with a single gap in the resonator.

[0039] FIGS. 19A-19B show examples of Raman spectra of proteins and protein fragments according to some embodiments.

[0040] FIGS. 20A-20C show examples of post translational modification spectra and associated confusion matrix, according to some embodiments.

[0041] FIG. 21 shows an example of a Raman emission versus excitation wavelength plot, according to some embodiments.

[0042] FIGS. 22A-22C show examples of fabricated arrays at different magnification levels, according to some embodiments.

[0043] FIGS. 23A-23C show far field scattering profiles of arrays, according to some embodiments.

[0044] FIGS. 24A-24B show examples of simulated clustered analysis of analytes and the associated simulated Raman spectra, according to some embodiments.

[0045] FIGS. 25A-25B show an example of a cluster analysis of a plurality of Raman spectra to identify analytes, according to some embodiments.

[0046] FIGS. 26A-26B show an example of Raman spectra and a difference spectrum associated with the introduction of a small molecule to a peptide, according to some embodiments.

[0047] FIGS. 27A-27D show an example of a confusion matrix and associated Raman spectra, according to some embodiments.

[0048] FIGS. 28A-28B show an example of a confusion matrix and associated Raman spectra, according to some embodiments.

[0049] FIG. 29 depicts an example of fingerprinting and sequencing a heterogenous sample.

[0050] FIG. 30 depicts an example of a process for identifying unknown proteins using a catalog of spectra.

[0051] FIG. 31 shows shows an example of a process for de novo peptide sequencing of unknown peptides.

[0052] FIG. 32 shows an example of how alignment marks are deposited.

[0053] FIG. 33 shows an example of how resonators are deposited.

[0054] FIG. 34 shows an example of how nanostructures arc deposited.

[0055] FIG. 35 shows the resonators across a field and the nanostructure located between resonators in a blown up section as well as the concentration of the E field around the nanostructure.

[0056] FIG. 36 shows the deposition of nanostructures between each set of resonators.

[0057] FIG. 37 shows a single nanostructure between a single set of resonators and the remaining sets of resonators absent a nanostructure.

[0058] FIG. 38 shows variation for the placement of metal with a nanostructure.

[0059] FIG. 39 shows variations for the placement of metal with a nanostructure relative to a resonator and / or an array.

[0060] FIG. 40 shows variations for the placement and shape of the metal nanostructure relative to the nanogap and the corresponding E field produced.

[0061] FIG. 41A shows an example of a quality factor Q for a hybrid silicon / metal resonator, according to some embodiments.

[0062] FIGS. 41B-41C show properties of a plurality of resonators comprising various amounts of nanostructures, according to some embodiments.

[0063] FIG. 42 shows an example of the backfill model for generating a functionalized nanostructure.

[0064] FIG. 43 shows an example of the co-deposition approach for generating a functionalized nanostructure.

[0065] FIG. 44 shows properties that a moiety may comprise when being used for generating a functionalized nanostructure.

[0066] FIG. 45 shows an example of processing a solution with a peptide and attaching it to a surface (e.g., surface of a nano structure or nanogap).

[0067] FIG. 46 shows an example of a plurality of functionalized nanostructures, according to some embodiments.

[0068] FIG. 47 shows a variety of examples if different surface chemistries to attach peptides to a surface.

[0069] FIGS. 48A-48B show exemplary resonance and surface field enhancement data. FIG. 48A shows dielectric and antennae resonances, for silicon and metal, respectively. FIG. 48B shows surface field enhancement from the antennae design.

[0070] FIGS. 49A-49B show design and simulation of a single-resonant high-Q resonance chip design, approximately. FIG. 49A shows the resonator orientation totaling ~8 um in length, which has a metal dot. FIG. 48B shows simulated |E|2 data, for a center cut line at peak resonance.

[0071] FIGS. 50A-50C depict the overall principle of resonance stacking for multi-resonant chip designs having hybrid dielectric and antennae designs (e.g., silicon and metal).

[0072] FIGS. 51A-5 IB depict a comparison of single high resonance vs. dual resonance antennae designs. FIG. 51 A shows single resonance structure with a 5 nm gap between features, and a dual resonance structure with a 20 nm gap between features. FIG. 5 IB shows Raman enhancement of the two approaches, indicating that the double resonance structures have enhanced signal over a range of minimum feature gap distances.

[0073] FIGS. 52A-52D depict unit cell schematics of a design (e.g., cichlid design) having dual antennae structures and resulting patterning of unit cells having a ratio of 1 : 1 primary to secondary photonic pillars. FIG. 52A shows the top down view of the primary photonic pillar with the antennae, and surrounding secondary photonic pillars. A unit cell may have whole or partial sections of secondary photonic pillars depending on the ratio of primary to secondary pillars and patterning on the chip. FIG. 52B shows a side view of the same design, emphasizing the manufacturing layering strategy (e g ., pillars on top of a substrate). FIGS. 52C-52D shows microscopic images of a manufactured cichlid chip at different zoom levels, where the scale bar is 1 um in FIG. 52C and 500 nm in FIG. 52D.

[0074] FIGS. 53A-53D demonstrate improved laser line tolerance of hybrid metal -di electric dual resonance chip designs. FIG. 53A shows the Raman intensity peak of a single high-Q only device around the laser w avelength, having sharp drop off center line. FIG. 53B-53C show improved Raman enhancement factor for the hybrid metal-dielectric dual resonance chip across broader laser wavelengths, both simulated (FIG. 53B) and experimentally (FIG. 53C). FIG. 53D shows Raman intensity counts of the hybrid chip for pump wavelengths of 1030 nm, 1030 nm, 1050 nm, and 1060 nm.

[0075] FIGS. 54A-54B demonstrate improved optical loss tolerance of hybrid metal-dielectric dual resonance chip designs. FIGS. 54A shows the Raman enhancement factor for a single high-Q device, and FIG. 54B shows the Raman enhancement factor for a hybrid resonance device.

[0076] FIGS. 55A-55B demonstrate improved incident light coupling tolerance of hybrid metaldielectric dual resonance chip designs. FIGS. 55 A shows that the Raman enhancement factor for a single high-Q device drops rapidly as a function of illumination beam angle, and FIG. 55B shows that the hybrid resonance device demonstrates off-angle Raman enhancement.

[0077] FIG. 56 shows a comparison between a discus design having a dual antennae (top) and cichlid designs having single and dual antennae designs (bottom).

[0078] FIG. 57A-57C show geometric variations in alignment of antennae versus centerline of a photonic pillar resonator. FIG. 57A shows placement of the antennae can vary in (x,y) placement in a primary photonic pillar. FIG. 57B shows vertical misalignments of antennae on a photonic pillar (0 nm, 20 nm, and 40 nm) and effects are shown in FIG. 57C.

[0079] FIGS. 58A-58C show geometric variations in lower layer photonic pillar and upper layer antennae measurement parameters. FIG. 58A shows results for varying photonic pillar disk radius. FIG. 58B shows results for varying the width of the unit cell. FIG. 58C shows results for varying the length of the antennae (e.g. bowtie length).

[0080] FIG. 59 shows geometric measurements of antennae features that are optimized based on laser and resonance considerations, including gap size, length, curvature, and thickness.

[0081] FIGS. 60A-60D show reflectance data while tuning photonic pillar disk radius (FIGS. 60A and 60B) and adjusting the unit cell height (FIGS. 60C and 60D).

[0082] FIGS. 61A-61F show validation excitation / emission data for optimizing photonic pillar and antennae dimensions with adjustments for varying pump wavelengths.

[0083] FIGS. 62A-62D show simulations of a cichlid chip designs to reduce crosstalk between photonic pillars at high sensor density.

[0084] FIGS. 63A-63F depict crosstalk optimization by rotating antennae on a photonic pillar. FIGS. 62A-62C show simulation of cross talk when antennae directions are aligned, and FIGS. 62D-62F show improvements when alternating antennae are rotated (shown is 90 degrees).

[0085] FIG. 64 depicts photonic unit cell design approaches having adjusted antennae design and placement orientation.

[0086] FIG. 65 depicts photonic pillar designs having polygonal shapes as an alternative to elliptical (e.g., circular) shapes. Mixed shapes, sizes, and rotational angles are shown.

[0087] FIG. 66 depicts unit cells having different ratios of primary to secondary photonic pillars.

[0088] FIG. 67 depicts a pattern having primary photonic pillars with different antennae designs (e.g., optimized for different wavelengths).

[0089] FIG. 68A-68D depict layering approaches when manufacturing chips using antennae and photonic pillars for optimized vibrational spectra.

[0090] FIG. 69A-68H depict performance improvements from mirror-containing chip designs.FIGS. 69A-69E show layering approaches when manufacturing chips to feature mirror surfaces below the antennae in a photonic pillar. FIGS. 69F-69H show the effect of varying the mirror layers.

[0091] FIGS. 70A-70Q show manufacturing lithography steps for producing chips for vibrational spectroscopy having photonic pillars, antennae, and mirror-surfaces.

[0092] FIG. 71 depicts a chip read-out process of a nanophotonic device designed around thermal influence, emission crosstalk, and field of view of the spectrometer.

[0093] FIG. 72 depicts a graphical overview of label-free vibrational spectroscopy base methods described herein (e.g., a Raman-based deSIPHR platform)

[0094] FIG. 73 depicts application of vibrational spectroscopy in a method for monitoring live cell interactions with a stimulant using barcodes (e.g, a Raman-based cellMAPP platform).

[0095] FIG. 74 depicts application of vibrational spectroscopy in a method for monitoring TCR recognition and unactivated, activated, and exhausted T cell states using barcodes.

[0096] FIG. 75 depicts Raman barcoding microbeads and attachment to DNA barcoded pMHC molecules for monitoring cell response to stimuli (e.g., antigens).

[0097] FIG. 76 depicts a comparison of the benefits vibrational spectroscopy compared to standard approaches for live cell monitoring.

[0098] FIG. 77 depicts the combination of vibrational spectroscopic cell monitoring methods with cleavable linkages and dye placements.

[0099] FIG. 78 depicts a workflow for capturing Raman fingerprints of unactivated and activated CD8+ T cells.

[0100] FIG. 79 depicts workflow for recovery of activated and exhausted CD8+ T cells for further processing (e.g., sequencing).

[0101] FIG. 80 depicts strategies for attaching Raman reporters directly to antigen presenting cells.FIG. 81 shows a comparison of the vibrational spectroscopy (e.g., Raman spectroscopy) proteomics workflow with traditional mass spectrometry.FIGS. 82A-82B depict graphical flow charts describing key steps in the vibrational spectroscopy (e.g., Raman spectroscopy) proteomics approach. FIG. 82C depicts a graphical flow chart of sample preparation.FIGS. 83A-83C depict chromatograms of peptides fractionation. FIG. 83A shows mobile phase solvent effects on peptide fractionation. FIG. 83B shows results following peptide sample doping, FIG. 83 C shows mobile phase gradient effects.FIG. 84 depicts molecules for chemical attachment of peptides to a surface.FIG. 85 depicts anchor options for chemical attachment of peptides to a surface.FIGS. 86A-86B depict chemical attachment strategies for peptides to surfaces. FIG. 86A shows an N-terminal approach for peptide immobilization via activated resins and release for surface conjugation. FIG. 86B shows effect of pH for release peptides from resin.FIG. 87 shows amino acid side chain chemistries, and the similarity of aspartic acid (D) and glutamic acid (E) with the C-terminus of a peptide.FIG. 88 depicts a schematic of specifying C-terminal modification and attachment of a peptide using oxazolone.FIG. 89 shows Raman spectra obtained for a bifunctional silane-azide linker (orange) after binding a peptide modified with an alkyne group via click chemistry (yellow)FIG. 90 shows Raman spectra obtained for bifunctional thiol-NHS ester linker (brown) and after binding of the peptide to the surface via the linker (orange) for surface anchoring to a metal surface.FIG. 91 shows Raman spectra obtained for a peptide attachment to a photonic chip having a cichlid design as described herein. Left side spectra shows baseline of a clean sensor before(brown top line) and following attachment of the peptide (bottom blue). The right three spectra show spectra obtained from three different devices areas of the device.FIG. 92 depicts automated liquid dispensers for micro-, nano-, and pico-well arrays, as well as flat and patterned substrates.FIG. 93A-93B depict advantages of small sample volumes. FIG. 93A shows that efficiencies in peptide concentrations occur when smaller volumes are used. FIG. 93B shows a microscopic image of rapidly dried wells.FIGS. 94A-94D depict droplet deposition into wells of differing designs. FIG. 94A shows dye- tagged peptide deposited on a commercially available well plate, with its Raman spectrum shown in FIG. 94B. FIG. 94C shows dye-tagged peptide deposited on a custom designed and well plate, with its Raman spectrum shown in FIG. 94D.FIGS. 95A-95B depict improved attachment following surface modification. FIG. 95A shows hydrophobic and omniphobic coatings can be places on surfaces for improved localization of attachment. The left side of FIG. 94B shows a microscopic image of a peptide printed on functionalized metal (gold) coated surface, and the right side shows the same printed on piranha cleaned, and plasma cleaned oxide-coated gold surface.FIGS. 96A-96B depict single molecule at a sensor location. FIG. 96A shows each sensor detecting a peptide, and FIG. 96B shows these peptides having different copy numbers.

[0102] FIGS. 97A-97B depict multiple molecules at a sensor location. FIG. 97A shows multiple peptides contributing to the spectra. FIG. 97B shows that demixing and spectral deconvolution of the spectra can be achieved for accurate prediction of peptide sequences.DETAILED DESCRIPTION

[0103] While various embodiments of the invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only.Numerous variations, changes, and substitutions may occur to those skilled in the art without departing from the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed.

[0104] Whenever the term “at least,” “greater than,” or “greater than or equal to” precedes the first numerical value in a series of two or more numerical values, the term “at least,” “greater than” or “greater than or equal to” applies to each of the numerical values in that series of numerical values. For example, greater than or equal to 1, 2, or 3 is equivalent to greater than or equal to 1, greater than or equal to 2, or greater than or equal to 3.

[0105] Wheneverthe tenu “no more than,” “less than,” or “less than or equal to” precedes the first numerical value in a series of two or more numerical values, the term “no more than,” “less than,” or “less than or equal to” applies to each of the numerical values in that series of numerical values. For example, less than or equal to 3. 2, or 1 is equivalent to less than or equal to 3, less than or equal to 2. or less than or equal to 1.

[0106] Spectroscopy, including Raman spectroscopy, spontaneous Raman spectroscopy, stimulated Raman spectroscopy (SRS), and coherent anti-stokes Raman spectroscopy (CARS) provides numerous benefits such as, for example, strong sensitivity to the composition of an analyte. Increased light intensity can provide higher signal, which can improve the ability to discern differences in spectra and detect features of analytes. Arrays of dielectric features with a nanogap in the features can significantly increase light field, thereby providing high signal.I. Systems for Raman Spectroscopy

[0107] In certain aspects, described herein are systems and methods for processing a biological sample. The systems and methods may comprise spectroscopy. The spectroscopy may comprise infrared (IR) spectroscopy, ultraviolet-visible (U V / Vis) spectroscopy, nuclear magnetic resonance (NMR) spectroscopy, Raman spectroscopy, X-ray spectroscopy, or any combination thereof. Hie spectroscopy may comprise Raman spectroscopy. The Ramen spectroscopy may be stimulated Raman spectroscopy (SRS). The Ramen spectroscopy may be coherent anti-Stokes Raman Spectroscopy (CARS). The Ramen spectroscopy may be spontaneous Raman spectroscopy. Raman spectroscopy provides numerous benefits such as, for example, strong sensitivity to the composition of an analyte. Increased light intensity can provide higher signal, which can improve the ability to discern differences in spectra and detect features of analytes. Arrays of dielectric features with a nanogap in the features can significantly increase light field, thereby providing high signal. The biological sample may comprise one or more components. In some embodiments, described herein is a chip for processing a biological sample. In some embodiments, the chip comprises an array of features. In some embodiments, the chip comprises a plurality of arrays of features. In some embodiments, the features are non-unifonn. In some embodiments, a feature comprises an electrical insulator or a semiconductor. In some embodiments, thefeature comprises a nanogap. In some embodiments, the array of features is interspersed with a plurality of electrodes or functionalized features configured to filter the one or more components according to size or charge. In some embodiments, the chip comprises or more resonators, wherein each of the two or more resonators supports one or more guided modes, wherein each of the two or more resonators has a corresponding longitudinal perturbation, wherein an incident light is coupled to two or more of the guided mode resonances by the longitudinal perturbations of the resonators, wherein each resonator comprises an electrical insulator or a semiconductor, wherein each resonator comprises at least one nanogap configured to concentrate an incident light, wherein one or more regions of high electromagnetic field intensity are localized within and in proximity to each nanogap, whereby environmental sensing is provided. In some embodiments, a chip comprises a resonator. In some embodiments, the resonator comprises a nanostructure. In some embodiments, the chip comprises a nanostructure.

[0108] As described herein, the terms “pump wavelength, excitation wavelength, and laser wavelength” may be used interchangeably. As described herein, the terms emission wavelength and scattered wavelength (e.g., Raman scattered) may be used interchangeably.(a) Chips

[0109] In certain aspects, described herein are chips comprising a plurality of features (e.g., a resonator comprising the array of non-uniform features). In some embodiments, the chips comprise a dielectric (e.g., silicon). The chip may comprise an array. The chip may comprise one or more arrays. An array may comprise two or more features. The features may be non-uniform. The features of the array may be parallel to one another. The features of the array may be nonparallel to each other. At least one feature of the array may be rectangular. At least one feature of the array may be rounded. The features of the array may be arranged in a periodic configuration. The features of the array may be arranged in a nonperiodic configuration. At least one feature of the array may comprise a guided mode resonance. At least one feature of the array may be a photonic crystal mirror. The array can be a resonator as described elsewhere herein. For example, the tenns array of non-uniform features and resonator can be used interchangeably.

[0110] The features described herein may comprise a nanogap. The features described herein may comprise a plurality of nanogaps. The nanogap and / or plurality of nanogaps may be configured to concentrate a light field coupled into the array. The nanogap and / or plurality of nanogaps may be formed as a part of the feature (e.g., formed at a same time as the feature). The nanogap and / or plurality of nanogaps may be formed after the feature (e.g., by removal of material from the feature). The features described herein may comprise a nanostructure. The features described herein may comprise a plurality of nanostructures. The nanostructure and / or plurality of nanostructures may be configured to concentratea light field coupled into a nanogap. The nanostructure may comprise a nanogap. The plurality of nanostructures may comprise a nanogap. The plurality of nanostructures may comprise a plurality of nanogaps. The features described herein may comprise a resonator. The features described herein may comprise a plurality of resonators. The resonator and / or plurality of resonators may be configured to concentrate a light field coupled into a nanogap. An array may comprise features, such as features as disclosed herein. In some embodiments, an array is a resonator.

[0111] An array may comprise a resonator. An array may be a resonator. A resonator may comprise a plurality of nanostructures. A resonator may comprise a plurality of nanostructures and a nanogap. A resonator may comprise a plurality of nanostructures and a plurality of nanogaps. A nanogap may be a vertical slit. A nanogap may be a vertical slit that goes down the length of a resonator. A nanogap may be the length of a nanostructure. A nanogap may be the length of a full resonator. A resonator may comprise about 0, about 1, about 2, about 3, about 4, about 5, about 6. about 7, about 8, about 9, about 10, about 11, about 12, about 13, about 14, about 15, about 16, about 17, about 18, about 19, about 20, about 25, about 30, about 35, about 40, about 45, about 50, about 55, about 60, about 65, about 70, about 75, about 80, about 85, about 90, about 95, about 100. about 105, about 110, about 115. about 120, about 125, about 130, about 135, about 140, about 145, about 150, about 155, about 160, about 165, about 170, about 175, about 180, about 185, about 190, about 195, about 200, about 205, about 210, about 215, about 220, about 225, about 230, about 235, about 240, about 245, about 250, about 255, about 260, about 265, about 270, about 275, about 280, about 285, about 290, about 295. about 300, about 305, about 310, about 315, about 320. about 325, about 330, about 335, about 340, about 345, about 350, about 355, about 360. about 365. about 370, about 375, about 380. about 385, about 390. about 395, about 400, about 405. about 410, about 415, about 420, about 425, about 430, about 435, about 440, about 445, about 450, about 455, about 460, about 465, about 470, about 475, about 480, about 485, about 490, about 495, about 500, or more nanogaps. A resonator may comprise at least about 0, about 1, about 2, about 3, about 4, about 5, about 6, about 7, about 8, about 9, about 10, about 11, about 12, about 13, about 14, about 15, about 16, about 17, about 18, about 19, about 20, about 25, about 30, about 35, about 40, about 45, about 50, about 55, about 60, about 65, about 70, about 75, about 80, about 85, about 90, about 95, about 100. about 105, about 110, about 1 15, about 120, about 125, about 130, about 135, about 140, about 145, about 150, about 155, about 160, about 165, about 170, about 175, about 180, about 185, about 190, about 195, about 200, about 205, about 210, about 215, about 220, about 225, about 230, about 235, about 240, about 245, about 250, about 255, about 260, about 265, about 270, about 275, about 280. about 285, about 290, about 295, about 300, about 305. about 310, about 315, about 320. about 325, about 330, about 335, about 340, about 345. about 350. about 355, about 360, about 365. about 370, about 375. about 380, about 385, about 390. about 395,about 400, about 405, about 410, about 415, about 420, about 425, about 430, about 435, about 440, about 445, about 450, about 455, about 460, about 465, about 470, about 475, about 480, about 485, about 490, about 495, about 500, or more nanogaps. A resonator may comprise greater than about 0, about 1, about 2, about 3, about 4, about 5, about 6, about 7, about 8, about 9, about 10, about 11, about 12, about 13, about 14, about 15, about 16, about 17, about 18, about 19, about 20, about 25, about 30, about 35, about 40, about 45, about 50, about 55, about 60, about 65, about 70, about 75, about 80, about 85, about 90, about 95, about 100, about 105, about 110, about 115, about 120, about 125, about 130, about 135, about 140, about 145, about 150, about 155, about 160, about 165, about 170, about 175, about 180, about 185, about 190, about 195, about 200, about 205, about 210, about 215, about 220, about 225, about 230, about 235, about 240, about 245, about 250, about 255, about 260. about 265, about 270, about 275, about 280, about 285. about 290, about 295, about 300, about 305, about 310, about 315, about 320, about 325. about 330. about 335, about 340, about 345. about 350, about 355. about 360, about 365, about 370. about 375, about 380, about 385, about 390, about 395, about 400, about 405, about 410, about 415, about 420, about 425, about 430, about 435, about 440, about 445, about 450, about 455, about 460, about 465, about 470, about 475, about 480, about 485, about 490, about 495, about 500, or more nanogaps.

[0112] A plurality of arrays may comprise a plurality of resonators. A plurality of arrays may comprise a plurality of nanostructures. A plurality of arrays may comprise a nanogap. A plurality of arrays may comprise a plurality of nanogaps. A plurality of arrays may comprise about 0, about 1, about 2, about 3, about 4, about 5, about 6, about 7, about 8, about 9, about 10, about 11, about 12, about 13, about 14, about 15, about 16, about 17, about 18, about 19, about 20, about 25, about 30, about 35, about 40, about 45, about 50, about 55, about 60, about 65, about 70, about 75, about 80, about 85, about 90, about 95, about 100, about 105, about 110, about 115, about 120, about 125, about 130, about 135, about 140, about 145, about 150, about 155, about 160, about 165, about 170, about 175, about 180, about 185, about 190, about 195, about 200, about 205, about 210, about 215, about 220, about 225, about 230, about 235, about 240, about 245, about 250, about 255, about 260, about 265. about 270, about 275, about 280, about 285, about 290. about 295, about 300, about 305. about 310, about 315, about 320, about 325, about 330. about 335. about 340, about 345, about 350. about 355, about 360. about 365, about 370, about 375. about 380, about 385, about 390, about 395, about 400, about 405, about 410, about 415, about 420, about 425, about 430, about 435, about 440, about 445, about 450, about 455, about 460, about 465, about 470, about 475, about 480, about 485, about 490, about 495, about 500, about 1,000, about 1,500, about 2,000, about2.500, about 3,000, about 3,500, about 4,000, about 4,500, about 5,000, about 5.500, about 6,000, about6.500, about 7,000, about 7,500. about 8,000, about 8.500, about 9,000, about 9.500, about 10,000, about 50,000, about 100,000, about 150.000, about 200.000. about 250,000. about 300,000, about 350,000,about 400,000, about 450,000, about 500,000, about 550,000, about 600,000, about 650,000, about 700,000, about 750,000, about 800,000, about 850,000, about 900,000, about 950,000, about 1 million (mil), about 2 mil, about 3 mil, about 4 mil, about 5 mil, about 6 mil, about 7 mil, about 8 mil, about 9 mil, about 10 mil, about 50 mil, about 100 mil, about 150 mil, about 200 mil, about 250 mil, about 300 mil, about 350 mil, about 400 mil. about 450 mil, about 500 mil, about 550 mil, about 600 mil, about 650 mil, about 700 mil, about 750 mil. about 800 mil, about 850 mil, about 900 mil. about 950 mil, about 1 bil (billion), about 2 bil, about 3 bil, about 4 bil, about 5 bil, about 6 bil, about 7 bil, about 8 bil, about 9 bil, about 10 bil, or more nanogaps. A plurality of arrays may comprise at least about 0, about 1, about 2, about 3, about 4, about 5, about 6, about 7, about 8, about 9, about 10, about 11, about 12, about 13, about 14, about 15, about 16, about 17, about 18, about 19, about 20, about 25, about 30, about 35, about 40, about 45, about 50, about 55, about 60, about 65, about 70, about 75, about 80, about 85, about 90, about 95, about 100. about 105, about 110, about 115. about 120, about 125. about 130, about 135, about 140. about 145, about 150, about 155, about 160, about 165, about 170, about 175, about 180, about 185, about 190, about 195, about 200, about 205, about 210, about 215, about 220, about 225, about 230, about 235, about 240, about 245, about 250, about 255, about 260, about 265, about 270, about 275, about 280, about 285, about 290, about 295, about 300, about 305, about 310. about 315, about 320, about 325, about 330, about 335. about 340, about 345, about 350. about 355, about 360, about 365, about 370, about 375. about 380. about 385, about 390, about 395. about 400, about 405. about 410, about 415, about 420. about 425, about 430, about 435, about 440, about 445, about 450, about 455, about 460, about 465, about 470, about 475, about 480, about 485, about 490, about 495, about 500, about 1,000, about 1,500, about 2,000, about2.500, about 3,000, about 3,500, about 4,000, about 4,500, about 5,000, about 5,500, about 6,000, about6.500, about 7,000, about 7,500, about 8,000, about 8,500, about 9,000, about 9.500, about 10,000, about 50,000, about 100,000, about 150.000, about 200,000. about 250,000, about 300,000, about 350,000, about 400.000, about 450.000. about 500.000. about 550,000. about 600,000, about 650,000, about 700,000, about 750,000, about 800,000, about 850,000, about 900,000, about 950,000, about 1 million (mil), about 2 mil, about 3 mil, about 4 mil, about 5 mil, about 6 mil, about 7 mil, about 8 mil, about 9 mil, about 10 mil, about 50 mil, about 100 mil, about 150 mil, about 200 mil, about 250 mil, about 300 mil, about 350 mil, about 400 mil, about 450 mil, about 500 mil, about 550 mil, about 600 mil, about 650 mil, about 700 mil, about 750 mil. about 800 mil, about 850 mil, about 900 mil, about 950 mil, about 1 bil (billion), about 2 bil, about 3 bil, about 4 bil, about 5 bil, about 6 bil, about 7 bil, about 8 bil, about 9 bil, about 10 bil, or more nanogaps. A plurality of arrays may comprise greater than about 0, about 1, about 2, about 3, about 4, about 5, about 6, about 7, about 8, about 9, about 10, about 11, about 12, about 13, about 14, about 15, about 16, about 17, about 18, about 19, about 20, about 25, about 30, about 35, about 40, about 45, about 50, about 55, about 60, about 65, about 70, about 75, about 80, about 85, about 90,about 95, about 100, about 105, about 110, about 115, about 120, about 125, about 130, about 135, about 140, about 145, about 150, about 155, about 160, about 165, about 170, about 175, about 180, about 185, about 190, about 195, about 200, about 205, about 210, about 215, about 220, about 225, about 230, about 235, about 240, about 245, about 250, about 255, about 260, about 265, about 270, about 275, about 280, about 285. about 290, about 295, about 300, about 305, about 310, about 315, about 320, about 325. about 330. about 335, about 340, about 345. about 350, about 355. about 360, about 365, about 370. about 375, about 380, about 385, about 390, about 395, about 400, about 405, about 410, about 415, about 420, about 425, about 430, about 435, about 440, about 445, about 450, about 455, about 460, about 465, about 470, about 475, about 480, about 485, about 490, about 495, about 500, about 1,000, about 1,500, about 2,000, about 2,500, about 3,000, about 3,500, about 4,000, about 4,500, about 5,000, about 5,500, about 6,000, about 6,500. about 7,000, about 7.500, about 8,000, about 8.500, about 9,000, about 9,500, about 10,000, about 50,000, about 100,000, about 150.000, about 200.000. about 250,000. about 300,000, about 350,000, about 400,000, about 450,000, about 500,000, about 550,000, about 600,000, about 650,000, about 700,000, about 750,000, about 800,000, about 850,000, about 900,000, about 950,000, about 1 million (mil), about 2 mil, about 3 mil, about 4 mil, about 5 mil, about 6 mil, about 7 mil, about 8 mil, about 9 mil, about 10 mil, about 50 mil, about 100 mil, about 150 mil, about 200 mil, about 250 mil, about 300 mil. about 350 mil, about 400 mil, about 450 mil, about 500 mil, about 550 mil, about 600 mil, about 650 mil. about 700 mil, about 750 mil, about 800 mil. about 850 mil, about 900 mil, about 950 mil, about 1 bil (billion), about 2 bil, about 3 bil, about 4 bil, about 5 bil, about 6 bil, about 7 bil, about 8 bil, about 9 bil, about 10 bil, or more nanogaps.

[0113] In some embodiments, at least one of the features of the array comprise a nanogap. In some embodiments, two or more features of the array comprise a nanogap. In some embodiments, three or more features of the array comprise a nanogap. In some embodiments, four or more features of the array comprise a nanogap. In some embodiments, five or more features of the array comprise a nanogap. In some embodiments, each of the features of tire array comprise a nanogap. In some embodiments, each feature may comprise about 0. 1, 2, 3, 4. 5, 6, 7, 8. 9, 10, or more nanogaps. In some embodiments, the features are non-uniform.

[0114] In some embodiments, the resonator comprising the nanogap is configured to concentrate an incident light. The incident light may be an incident laser, a light emitting diode (LED) light, a lamp, or a combination thereof. The incident light may be an LED. The incident light may be a lamp. In some embodiments, a light source of a first light is integrated with the chip. In some embodiments, a light source of a first light is not integrated with the chip. In some embodiments, a light source of a secondlight is integrated with the chip. In some embodiments, a light source of a second light is not integrated with the chip.

[0115] The incident light may be an incident laser. The incident laser may have a wavelength of at least at least about 100 nanometers (nm), at least about 200 nm, at least about 300 nm, at least about 400 nm, at least about 500 nm, at least about 600 nm, at least about 700 nm, at least about 800 nm, at least about 900 nm, at least about 1000 nm, at least about 1100 nm, at least about 1200 nm, at least about 1300 nm, at least about 1400 nm, at least about 1500 nm, at least about 1600 nm, at least about 1700 nm, at least about 1800 nm, at least about 1900 nm, at least about or 2000 nm. The incident laser may have a wavelength of no more than about 100 nm. about 200 nm, about 300 nm. about 400 nm, about 500 nm, about 600 nm, about 700 nm, about 800 nm, about 900 nm, about 1000 nm, about 1100 nm, about 1200 nm, about 1300 nm, about 1400 nm, about 1500 nm, about 1600 nm, about 1700 nm, about 1800 nm, about 1900 nm, about or 2000 nm. The incident laser may have a wavelength of at least about 400 nm to at least about 1800 nm. The incident laser may have a wavelength of at least about 100 nm to at least about 1000 nm. The incident laser may have a wavelength of at least about 300 nm to about 1800 nm, about 400 nm to about 1800 nm, about 500 nm to about 1800 nm, about 600 nm to about 1800 nm. about 700 nm to about 1800 nm, about 800 nm to about 1800 nm, about 900 nm to about 1800 nm, about 1000 nm to about 1800 nm, about 1100 nm to about 1800 nm, about 1200 nm to about 1800 nm, about 1300 nm to about 1800 nm, about 1400 nm to about 1800 nm, about 1500 nm to about 1800 nm, about 1600 nm to about 1800 nm or about 1700 nm to about 1800 nm. In some cases, the incident light can be generated by a plurality of light sources (e.g., lasers). For example, the incident light can be amixture of light from two lasers. In another example, the incident light can be light from a first laser and subsequently light from a second laser. The use of a plurality of light sources can enable use of pump-probe type excitation or detection schemes (e.g., stimulated Raman scattering, coherent anti-Stokes Raman, etc.).

[0116] The nanogap may comprise a binding moiety that is specific for an analyte. The nanogap may comprise a binding moiety that is non-specific for an analyte. In some embodiments, the nanogap comprises a binding moiety with binding specificity for said analyte. Tire binding moiety may be attached to a nanostructure. For example, the binding moiety can be linked to the surface of the nanostructure as described elsewhere herein (e.g., a thiol can bind to the surface of a gold nanostructure). The analyte may comprise a protein. The analyte may comprise an enzyme. The analyte may comprise a kinase. The analyte may comprise a receptor. The analyte may comprise a peptide. The analyte may comprise an antibody. The analyte may comprise a nucleic acid. The analyte may comprise nucleic acids. The analyte may comprise a small molecule. The analyte may comprise a metabolite (e.g., ametabolized derivative of a molecule). The analyte may comprise a tyrosine kinase. The analyte may comprise a Janus kinase 3. The analyte may comprise an epidermal growth factor receptor (EGFR).

[0117] The nanogap may be at least about 1 nm, about 2 nm, about 3 nm. about 4 nm, about 5 nm, about 6 nm, about 7 nm, about 8 nm, about 9 nm, about 10 nm, about 15 nm, about 20 nm, about 25 mu, about 30 nm, about 35nm, about 40 nm, about 45 nm, about 50 nm, about 60nm, about 70 nm, about 80 nm, about 90 nm, about lOOnm, about 110 nm, about 120 nm, about 130 nm, about 140 nm, about 150 nm, about 160 nm, about 170 nm, about 180 nm, about 190 nm, about or about 200 nm wide. Tire nanogap may be no more than about 2 nm, about 3 rnn, about 4 nm, about 5 nm, about 6 nm, about 7 nm. about 8 nm, about 9 nm. about 10 nm. about 15 nm, about 20 nm, about 25 nm, about 30 nm, about 35nm, about 40 nm, about 45 nm, about 50 nm, about 60nm, about 70 nm, about 80 nm, about 90 nm, about lOOnm, about 110 nm, about 120 nm, about 130 nm, about 140 nm, about 150 nm, about 160 nm, about 170 nm, about 180 nm, about 190 nm, about or about 200 nm wide. Tire nanogap may be at least about 5 nm to at least about 200 nm, about at least about 5 nm to at least about 150 nm, about at least about 5 nm to at least about 100 nm, about at least about 5 nm to at least about 50 nm. Tire nanogap may be at least about 5 nm to at least about 150 nm. In some embodiments, the nanogap comprises space for a single molecule to fit. In some cases, the nanogap comprises space for at most about 500, 400, 300, 200, 100, 90, 80, 70, 60, 50, 40, 30, 20, 10, 9, 8, 7, 6, 5, 4, 3, 2, or fewer molecules to fit. For example, the nanogap can be configured to have space between the nanostructure and the resonator less than the exclusion volume of two analytes of interest. In this example, a single analyte can bind to the nanostructure, and additional analytes can be excluded from the nanogap of the resonator by the combination of the volume of the nanostructure and the exclusion volume of the bound analyte. In some embodiments, the nanogap comprises a nanostructure. The nanogap may comprise space for the nanostructure and a single molecule (e.g., single analyte). Tire nanogap may not comprise space for the nanostructure and two or more molecules. Tire nanostructure may have a size of at most about 1,000, 900, 800, 700, 600, 500, 400, 300, 200, 100, 90, 80, 70, 60, 50, 40, 30, 20, 10, or fewer nanometers.

[0118] In certain aspects, the chips described herein comprise a resonator as described elsewhere herein. The resonator may comprise a nanostructure. The resonator may comprise a metallic nanostructure. Tire nanostructure may be disposed in a gap, such as a nanogap. lire nanostructure may be atop a gap, such as a nanogap. The nanogap may be any of the nanogaps disclosed herein. Tire nanostructure may be metallic, dielectric, or the like, or any combination thereof. Tire resonator may comprise features, such as the non-uniform features disclosed herein. The non-uniform features may be oriented to form a gap. The non-uniform features may comprise a gap. The resonator may comprise two or more elements. The two or more elements may be oriented to form a gap. Tire two or more elements may be separated by a gap.The two or more elements may comprise a gap. As a non-limiting example, the resonator may comprise a first element and a second element that are positioned in such a way as to form a gap. Alternatively, as another non-limiting example, the resonator may comprise a first element and a second element that are separated by a gap.

[0119] In some aspects, the chips described herein can comprise one or more pixels. In some embodiments, the one or more pixels are guided-mode resonance meta surface pixels, or GMR pixels. In some embodiments, the GMR pixels comprise a cavity. In some embodiments, the cavity comprises a resonator. In some embodiments, the GMR pixels comprise a gap, such as a nanogap, such as the nanogap previously disclosed herein. In some embodiments, the cavity comprises a gap, such as a nanogap, such as the nanogap disclosed elsewhere herein. In some embodiments, the gap comprises a nanostructure. In some embodiments, the cavity comprises a nanostructure. The nanostructure may be metallic. The nanostructure may be functionalized.

[0120] In some aspects, tire chips described herein comprise an antenna structure (e.g., a structure configured to localize at least a portion of an incident light). In some embodiments, the antenna structure comprises a gap, such as a nanogap, such as the nanogap previously disclosed herein. In some embodiments, the gap comprises a nanostructure. The nanostructure may be metallic. The nanostructure may be functionalized.

[0121] In some embodiments, the metallic nanostructure comprises gold. The metallic nanostructure may comprise platinum. The metallic nanostructure may comprise chromium. Tire metallic nanostructure may comprise titanium. The nanostructure may be an alloy. The Non-limiting examples of alloys include mixtures of gold and chromium, mixtures of gold and titanium, mixtures of chromium and platinum, and mixtures of gold and platinum.

[0122] The nanostructure may be functionalized. The nanostructure may be a functionalized nanostructure. The nanostructure or functionalized nanostructure may be chemically functionalized. The chemical functionalization may comprise reducing a density of a tethering molecule in a non-specific binding reducing agent. The reducing of the density of a tethering molecule in a non-specific binding reducing agent may comprise backfilling into an established matrix (FIG. 42). The reducing of the density of a tethering molecule in a non-specific binding reducing agent may comprise backfilling into an established surface. Tire reducing of the density of a tethering molecule in a nonspecific binding reducing agent may comprise co-depositing with a matrix (FIG. 43). Tire reducing of the density of a tethering molecule in a non-specific binding reducing agent may comprise co-depositing with a surface. Thereducing of the density of a tethering molecule in a non-specific binding reducing agent may comprise codepositing with the agent. The binding reducing agent may be configured to fill at least a portion of the surface of the nanostructure in order to passivate the surface of the nanostructure and reduce binding to the bare surface. For example, a non-functionalized alkyl thiol can be deposited onto a gold nanostructure to passivate the surface of the nanostructure. The binding reducing agent can be configured with one or more moieties that do not bind to analytes or other constituents of a solution flowed over the nanostructure. For example, the binding reducing agent can be configured not to bind to any of the constituents of a sample, thereby reducing the amount of surface on the nanostructure that is available to bind to.

[0123] The nanostructure or functionalized nanostructure may be functionalized with one or more moieties. Tire one or more moieties may be able to reduce non-specific binding. The one or more moieties may be able to reduce non-specific binding by using a matrix filled with the one or more moieties. Tire one or more moieties may be able to reduce non-specific binding by using a surface filled with the one or more moieties. The one or more moieties may be able to reduce non-specific binding byusing an agent filled with the one or more moieties. The one or more moieties may include one or more different moieties. The one or more moieties may comprise molecules comprising oxygen groups, alkyl groups, zwitterionic groups, or any combination thereof. The one or more moieties may comprise alkylthiols, alkyl-oxysilanes, alkyl-selenols, oxygen containing alkyl-thiols, oxygen containing alkyloxysilanes, oxygen containing alkyl-selenols, zwitterionic containing thiols, zwitterionic containing oxysilanes, zwitterionic containing selenols. cyclic containing thiols, cyclic containing oxysilanes, cyclic containing selenols, azide containing thiols, azide containing oxysilanes, azide containing selenols, alkyne containing thiols, alkyne containing oxysilanes, alkyne containing selenols, or any combination thereof. The one or more molecules may comprise one or more of alkyl-thiols, alkyl-oxysilanes, alkyl-selenols, oxygen containing alkyl-thiols, oxygen containing alkyl-oxysilanes, oxygen containing alkyl-selenols, zwitterionic containing thiols, zwiterionic containing oxysilanes, zwiterionic containing selenols, cyclic containing thiols, cyclic containing oxysilanes, cyclic containing selenols. azide containing thiols, azide containing oxysilanes, azide containing selenols, alkyne containing thiols, alkyne containing oxysilanes, alkyne containing selenols, or any combination thereof (FIG. 44). FIG. 46 shows an example of a plurality of functionalized nanostructures, according to some embodiments. The functionalization may be functionalization as described elsewhere herein. The functionalization can provide Poisson loading statistics as shown in FIG. 46. In some embodiments, the moieties may be click moieties. Non-limiting examples of click moieties include as functionalized alkynes, fluorophore alkynes and / or azides,nucleoside azides and / or alkynes, carbohydrate azides, organic azides, PEG azides, azide sources, amino acid azides and / or alkynes.

[0124] Nanostructures as disclosed herein may couple to a single molecule. The single molecule may be a biological molecule. A molecule may be, for example, a peptide, a protein, a small molecule, a metabolite, nucleic acids, or the like, or any combination thereof. The nanostructure may couple to a molecule through amide coupling. The amide coupling may be performed using a carbodiimide. The amide coupling may be performed using an oxime. Non-limiting examples of reagents used for amide coupling include but are not limited to EDC (l-ethyl-3-(3'-dimethylaminopropyl)carbodiimide, DCC (dicyclohexylcarbodiimide). DIC (diisopropylcarbodiimide), NITS (N -hydroxysuccinimide), and oxyma (ethyl cyano(hydroxyimino)acetate). The nanostructure may couple to a molecule using thiol attachment. A molecule comprising cysteine may attach to a metallic nanostructure through direct attachment of a cysteine (Cys) to the metallic nanostructure. A long-chain thiol in solution may couple to a molecule. A long-chain thiol may couple to a molecule through solid-phase peptide synthesis. In these depicted embodiments, the long chain thiol in solution may also attach directly to the metallic nanostructure through a thiol-metal bond. In some embodiments, a thiol attachment may occur with a surface modified by maleimide. In some embodiments, a thiol attachment may occur with a nanostructure modified by maleimide. A peptide from a sample may be coupled to the nanostructure by altering the peptide’s C- terminus in solution. In some embodiments, the C-terminus is modified using alkynes. In some embodiments, the C-terminus is modified using azides. In some embodiments, the C-terminus is modified using amino thiol.

[0125] In some embodiments, a whole or fragmented molecule may be attached to any nanostructure as disclosed herein in a solution. In this depicted embodiment, tire whole or fragmented molecules are drop cast onto a surface comprising a nanostructure. In some embodiments, a whole or fragmented molecule is attached to tire nanostructure using free amines and epoxides. A free amine may comprise lysine. A free amine may comprise an N-terminus amine. Although several free amines are listed here, any free amine on the whole or fragmented molecule would be sufficient. In some embodiments, the epoxides are on the nanostructure. An epoxide may comprise 3 -glycidoxypropyltrimethoxy silane. Although an epoxide is listed here, any epoxide on the nanostructure would be sufficient. In some embodiments, a whole or fragmented molecule may comprise a cysteine. In this depicted embodiment, a free thiol in the whole or fragmented molecule would be able to directly attach to the metallic nanostructure.

[0126] In some embodiments, a molecule may be attached to any nanostructure disclosed herein using an organic solvent. Non-limiting examples of organic solvents include dimethylsulfoxide (DMSO) anddimethylformamide (DMF). Although several organic solvents are listed here, any organic solvent would be sufficient. In some embodiments, a molecule may be attached to a nanostructure using a solubilizing agent. In some embodiments, a molecule may be attached to a nanostructure using a detergent. Nonlimiting examples of solubilizing agents and detergents include hexafluoroisopropanol (HFIP), sodium lauryl sulphates, and polyethylene glycols (PEGs). Although several solubilizing agents and detergents are listed here, any solubilizing agents and detergents would be sufficient. In some embodiments, a molecule may be attached to a nanostructure using a solubilizing linker. A non-limiting example of a solubilizing linker includes poly-arginine. Although a solubilizing linker is listed here, any solubilizing linkers would be sufficient. A molecule may be attached to a nanostructure through crosslinker attachment. A crosslinker may be DSS (disuccinimidyl suberate). A crosslinker may be BS3 (bis[sulfosuccinimidyl] suberate). Although several crosslinkers are listed here, any crosslinkers would be sufficient.

[0127] In some embodiments, the chip comprises an additional array comprising one or more features. The features may be non-uniform. The features may be configured to filter two or more components of the biological sample according to size, charge, or binding affinity. In some embodiments, the additional array is configured to filter the two or more components according to size. In some embodiments, the additional array is configured to filter the two or more components according to charge. In some embodiments, the additional array is configured to filter the two or more components according to binding affinity. In some embodiments, the non-uniform features of said additional array are interspersed with a plurality of electrodes or functionalized features configured to filter the two or more components according to charge or size. In some embodiments, the functionalized feature comprises a functionalized oxide surface. In some embodiments, the array of non-uniform features is interspersed with a plurality of electrodes or functionalized features configured to filter the two or more components according to size or charge. In some embodiments, the biological sample comprises at least about 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 150, 200, 250, 300, 350, 400, 450, 500, or more components.

[0128] In some embodiments, a feature of the array comprises an electrical insulator or a semiconductor. In some embodiments, the feature comprises one or more materials from the group consisting of silicon, silicon nitride, aluminum nitride, titanium dioxide, silicon dioxide, gallium nitride, hafnium oxide, germanium, gold, platinum, chromium, and silicon carbide. In some embodiments, the chip is provided on a substrate. In some embodiments, the substrate comprises one or more materials from the group consisting of germanium aluminum oxide, silicon dioxide, fused silica, silicon dioxide on silicon, silicon, silicon nitride, gallium nitride, calcium fluoride, and beryllium oxide.

[0129] At least one feature of the array described herein may have a height of at least about 10 nm, about 20 nm, about 30 nm, about 40 nm, about 50 nm, about 60 nm, about 70 nm, about 80 nm, about 90 nm, about 100 nm, about 110 nm, about 120 nm, about 130 nm, about 140 nm, about 150 nm, about 160 nm, about 170 nm, about 180 nm, about 190 nm. about 200 nm, about 250 nm, about 300 nm, about 350 nm, about 400 nm. about 450 nm, about 500 nm. about 550 nm, about 600 nm, about 650 nm, about 700 nm, about 750 nm, about 800 nm, about 850 nm, about 900 nm, about 950 nm, about 1000 nm, about 1100 nm, about 1200 nm, about 1300 nm, about 1400 nm, about 1500 nm, about 1600 nm, about 1700 nm, about 1800nm, about 1900 nm, about or about 2000 nm. In some embodiments, the height is no more than about 20 mn, about 30 nm, about 40 nm, about 50 nm, about 60nm, about 70 nm, about 80 nm, about 90 nm, about lOOnm, about 110 nm. about 120 nm, about 130 nm, about 140 nm, about 150 nm, about 160 nm. about 170 nm, about 180 nm. about 190 nm, about 200 nm. about 250 nm, about 300 mn, about 350 nm, about 400 nm, about 450 nm, about 500 nm, about 550 nm, about 600 nm, about 650 nm, about 700 nm, about 750 nm, about 800 nm, about 850 nm, about 900 mn, about 950 nm, about 1000 nm, about 1100 nm, about 1200 nm, about 1300 nm, about 1400 nm, about 1500 nm, about 1600 nm, about 1700 nm, about 1800nm, about 1900 nm, about or 2000 nm. In some embodiments, the height is at least about 10 nm to at least about 1000 nm. In some embodiments, the height is at least about 20 nm to at least about 1000 nm. In some embodiments, the height is at least about 30 nm to at least about 1000 nm. In some embodiments, the height is at least about 40 nm to at least about 1000 nm. In some embodiments, the height is at least about 10 nm to at least about 1000 nm. In some embodiments, the height is at least about 50 nm to at least about 1000 nm. In some embodiments, the height is at least about 60 nm to at least about 1000 nm. In some embodiments, the height is at least about 70 nm to at least about 1000 nm. In some embodiments, the height is at least about 80 nm to at least about 1000 nm. In some embodiments, the height is at least about 90 nm to at least about 1000 nm. In some embodiments, the height is at least about 100 nm to at least about 1000 nm.

[0130] At least one feature of the array described herein may have a width of at least about 10 nm, about 20 nm, about 30 nm, about 40 nm. about 50 nm. about 60 nm. about 70 nm. about 80 nm, about 90 nm, about 100 nm, about 110 nm, about 120 nm, about 130 nm, about 140 nm, about 150 nm, about 160 nm, about 170 nm, about 180 nm, about 190 nm, about 200 nm, about 250 nm, about 300 nm, about 350 nm, about 400 nm, about 450 nm, about 500 nm, about 550 nm, about 600 nm, about 650 nm, about 700 nm, about 750 nm, about 800 nm, about 850 nm. about 900 nm, about 950 nm, about 1000 nm, about 1100 nm, about 1200 nm, about 1300 nm. about 1400 nm, about 1500 nm, about 1600 nm, about 1700 nm, about 1800nm, about 1900 nm, about or 2000 nm. In some embodiments, the width is no more than about 20 nm, about 30 nm, about 40 nm, about 50 nm, about 60nm, about 70 nm, about 80 nm, about 90nm, about lOOnm, about 110 nm, about 120 nm, about 130 nm, about 140 nm, about 150 nm, about 160 nm, about 170 nm, about 180 nm, about 190 nm, about 200 nm, about 250 nm, about 300 nm, about 350 nm, about 400 nm, about 450 nm, about 500 nm, about 550 nm, about 600 nm, about 650 nm, about 700 nm, about 750 nm, about 800 nm, about 850 nm, about 900 nm, about 950 nm. about 1000 nm, about 1100 nm, about 1200 nm, about 1300 nm, about 1400 nm. about 1500 nm, about 1600 nm, about 1700 nm, about 1800nm, about 1900 nm, about or 2000 nm. In some embodiments, the width is at least about 10 nm to at least about 500 nm. In some embodiments, the width is at least about 50 nm to at least about 500 nm. In some embodiments, the width is at least about 50 nm to at least about 1000 nm.

[0131] At least one feature of the array described herein may have a width of at least about 10 nm, about 20 nm, about 30 nm, about 40 nm, about 50 nm, about 60 nm, about 70 nm, about 80 nm, about 90 nm, about 100 nm, about 110 nm, about 120 nm, about 130 nm, about 140 nm, about 150 nm, about 160 nm, about 170 nm, about 180 nm, about 190 nm, about 200 nm, about 250 nm, about 300 nm, about 350 nm, about 400 nm, about 450 nm, about 500 nm. about 550 nm, about 600 nm, about 650 nm, about 700 nm, about 750 nm. about 800 nm, about 850 nm. about 900 nm, about 950 nm, about 1000 nm, about 1100 nm, about 1200 nm. about 1300 nm. about 1400 nm, about 1500 nm, about 1600 nm, about 1700 nm, about 1800nm, about 1900 nm, about or 2000 nm. In some embodiments, the width is no more than about 20 nm, about 30 nm, about 40 nm, about 50 nm, about 60nm, about 70 nm, about 80 nm, about 90 nm, about lOOmn, about 110 nm, about 120 nm, about 130 nm, about 140 nm, about 150 nm, about 160 nm, about 170 nm, about 180 nm, about 190 nm, about 200 nm, about 250 nm, about 300 nm, about 350 nm, about 400 nm, about 450 nm, about 500 nm, about 550 nm, about 600 nm, about 650 nm, about 700 nm, about 750 nm, about 800 nm, about 850 nm, about 900 nm. about 950 nm, about 1000 nm, about 1100 nm, about 1200 nm, about 1300 nm, about 1400 nm, about 1500 nm, about 1600 nm, about 1700 nm, about 1800nm, about 1900 nm, about or 2000 nm. In some embodiments, the length is at least about 50 nm to at least about 2000 nm. In some embodiments, the length is at least about 100 nm to at least about 2000 nm. In some embodiments, the length is at least about 200 nm to at least about 2000 nm. In some embodiments, the length is at least about 300 nm to at least about 2000 nm. In some embodiments, the length is at least about 400 nm to at least about 2000 nm. In some embodiments, the length is at least about 500 nm to at least about 2000 nm. In some embodiments, the length is at least about 600 nm to at least about 2000 nm. In some embodiments, the length is at least about 700 nm to at least about 2000 nm. In some embodiments, the length is at least about 800 nm to at least about 2000 nm. In some embodiments, the length is at least about 900 nm to at least about 2000 nm. In some embodiments, the length is at least about 1000 nm to at least about 2000 nm.

[0132] The features described herein may be separated. In some embodiments, the distance between the features is at least about 10 nm, at least about 20 nm, at least about 30 nm, at least about 40 nm, at least about 50 nm, at least about 60 nm, at least about 70 nm, at least about 80 nm, at least about 90 nm, at least about 100 nm, at least about 110 nm, at least about 120 nm, at least about 130 nm, at least about 140 nm, at least about 150 nm, at least about 160 nm, at least about 170 nm, at least about 180 nm. at least about 190 nm, at least about 200 nm, at least about 250 nm, at least about 300 nm, at least about 350 nm, at least about 400 nm, at least about 450 nm, at least about 500 nm, at least about 550 nm, at least about 600 nm, at least about 650 nm, at least about 700 nm, at least about 750 nm, at least about 800 nm, at least about 850 nm, at least about 900 nm, at least about 950 nm, at least about 1000 nm, at least about 1100 nm, at least about 1200 nm, at least about 1300 nm, at least about 1400 nm, at least about 1500 nm, at least about 1600 nm, at least about 1700 nm, at least about 1800nm. at least about 1900 nm. or at least about 2000 nm. In some embodiments, the distance between the features is no more than about 20 nm, about 30 nm, about 40 nm, about 50 nm, about 60 nm, about 70 nm, about 80 nm, about 90 nm, about lOOnni, about 110 nm, about 120 nm, about 130 nm, about 140 nm, about 150 nm, about 160 nm, about 170 nm, about 180 nm, about 190 nm, about 200 nm, about 250 nm, about 300 nm, about 350 nm, about 400 nm, about 450 nm, about 500 nm. about 550 nm, about 600 nm. about 650 nm, about 700 nm, about 750 nm. about 800 nm, about 850 nm. about 900 nm, about 950 nm. about 1000 nm, about 1100 nm, about 1200 nm, about 1300 nm, about 1400 nm, about 1500 nm, about 1600 nm, about 1700 nm, about 1800nm, about 1900 nm, about or 2000 nm. In some embodiments, the distance is at least about 10 nm to at least about 1000 nm. In some embodiments, the distance is at least about 20 nm to at least about 1000 nm. In some embodiments, tire distance is at least about 30 nm to at least about 1000 nm. In some embodiments, the distance is at least about 40 nm to at least about 1000 nm. In some embodiments, the distance is at least about 10 nm to at least about 1000 nm. In some embodiments, the distance is at least about 50 nm to at least about 1000 nm. In some embodiments, the distance is at least about 60 nm to at least about 1000 nm. In some embodiments, the distance is at least about 70 nm to at least about 1000 nm. In some embodiments, the distance is at least about 80 nm to at least about 1000 nm. In some embodiments, the distance is at least about 90 nm to at least about 1000 nm. In some embodiments, the distance is at least about 100 nm to at least about 1000 nm. In some embodiments, the features are non- uniform.

[0133] Two or more of tire features of the array may be parallel to one another. Two or more of the features of the array may be non-parallel to each other.

[0134] Tire chip may comprise a first subset of an array of features and a second subset of an array of features. Tire first subset and the second subset may be adjacent to each other. Tire first subset and thesecond subset may be parallel to each other. The first subset may be parallel to the second subset. In some embodiments, the first subset is not parallel to the second subset. In some embodiments, the first subset is separated from the second subset by one or more dielectric fins.

[0135] The first subset and the second subset may be separated by a distance of at least about 1 nm, about 2 nm, about 3 nm, about 4 nm, about 5 nm, about 6 nm, about 7 nm, about 8 nm, about 9 nm, about 10 nm, about 15 nm, about 20 nm, about 25 nm, about 30 nm, about 35nm, about 40 nm, about 45 nm, about 50 nm, about 60nm, about 70 mn, about 80 nm, about 90 nm, about lOOnm, about 110 nm, about 120 nm, about 130 nm, about 140 nm. about 150 nm, about 160 nm. about 170 nm, about 180 mn, about190 nm. about 200 nm, about 250 nm. about 300 nm, about 350 nm. about 400 nm, about 450 mn, about500 nm, about 550 nm, about 600 nm, about 650 nm, about 700 nm, about 750 nm, about 800 nm, about850 nm, about 900 nm, about 950 nm, about 1000 nm, about 1100 nm, about 1200 nm, about 1300 nm, about 1400 nm, about 1500 nm, about 1600 nm, about 1700 nm, about 1800nm, about 1900 nm, about 2000 nm, about 2100 nm, about 2200 nm, about 2300 nm, about 2400 nm, about 2500 nm, about 2600 nm, about 2700 nm, about 2800 nm. about 2900 nm, about or 3000 nm. The first subset and the second subset may be separated by a distance of no more than about 5 nm, about 6 nm, about 7 nm. about 8 nm, about 9 nm, about 10 nm, about 15 nm, about 20 nm, about 25 nm, about 30 nm, about 35nm, about 40 nm, about 45 nm, about 50 nm, about 60nm, about 70 nm, about 80 nm, about 90 nm, about lOOnm, about 110 nm, about 120 nm, about 130 nm, about 140 nm, about 150 nm, about 160 nm, about 170 nm, about180 nm, about 190 nm, about 200 nm, about 250 nm, about 300 nm, about 350 nm, about 400 nm, about450 nm, about 500 nm, about 550 nm. about 600 nm, about 650 nm. about 700 nm, about 750 mn, about800 nm. about 850 nm, about 900 nm. about 950 nm, about 1000 nm, about 1100 nm. about 1200 nm. about 1300 nm, about 1400 nm, about 1500 nm, about 1600 nm, about 1700 nm, about 1800nm, about 1900 nm, about 2000 nm, about 2100 nm, about 2200 nm, about 2300 nm, about 2400 nm, about 2500 nm, about 2600 nm, about 2700 mn, about 2800 nm, about 2900 nm, about or 3000 nm. The distance may be at least about 3000 nm. Tire distance may be at least about 2000 nm. Tire distance may be at least about 1000 mn. The distance may be at least about 5 nm to at least about 3000 nm. The distance may be at least about 5 nm to at least about 2500 nm. The distance may be at least about 5 nm to at least about 2000 nm. The distance may be at least about 5 nm to at least about 1000 nm.

[0136] In some embodiments, the chips described herein comprise two or more resonators. In some embodiments, each of the two or more resonators supports one or more guided modes. In some embodiments, each of the two or more resonators has a corresponding longitudinal perturbation, wherein an incident light is coupled to two or more of the guided mode resonances by the longitudinal perturbations of the resonators. In some embodiments, each resonator comprises an electrical insulator ora semiconductor. In some embodiments, each resonator comprises at least one nanogap configured to concentrate an incident light as described herein. In some embodiments, each resonator comprises about 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 nanogaps. In some embodiments, one or more regions of high electromagnetic field intensity are localized within and in proximity to each nanogap, whereby environmental sensing is provided. In some embodiments, a resonator comprises a binding moiety that is non-specific for an analyte described herein. In some embodiments, a resonator comprises a binding moiety that is specific for an analyte described herein. In some embodiments, the resonator does not comprise a binding moiety. In some embodiments, two or more resonators of the array are parallel to one another. In some embodiments, two or more resonators of the array are nonparallel to one another. In some embodiments, each resonator further comprises a photonic crystal mirror designed to confine the guided mode resonance.

[0137] An array can have a quality factor (Q) descriptive of the efficiency of the array at concentrating electric field. A higher Q can be related to an array that more efficiently concentrates an incident light field, resulting in higher field strengths in the array. The arrays of the present disclosure can have a quality factor of at least about 100, 200. 300, 400. 500, 600. 700, 800. 900, 1,000. 2,000, 3,000. 4.000, 5,000, 7,500, 10,000, 15,000, 20,000, 25,000, 35,000, 50,000, 75,000, 100,000, or more. The arrays of the present disclosure can have a quality factor of at most about 100,000, 75,000, 50,000, 35,000, 25,000, 20,000, 15,000, 10,000, 7,500, 5,000, 4,000, 3,000, 2,000, 1,000, 900, 800, 700, 600, 500, 400, 300, 200, 100, or less. The arrays of the present disclosure can have a quality factor in a range as defined by any two of the preceding values. Such Q can be achieved in part due to the presence of the gap in the array, which can result in the enhanced field localization that causes the high Q of the array. Similarly, the reduced volume of the gap can result in reduced mode volume of the resonator. The mode volume of the resonator can be less than the wavelength of the light used to excite the resonator. The mode volume of the resonator may be at most about 2,000, 1,900, 1,800, 1,700, 1,600, 1,500, 1,400, 1,300, 1,200, 1,100, 1,000, 900, 800, 700, 600, 500, or fewer nanometers. Examples of high Q resonators and the calculations related to such resonators can be found in “Very-Large-Scale Integrated High-Q Nanoantenna Pixels ( VINPix)” by Varun Dolia et. al.. arXiv preprint arXiv:2310.08065 (2023). which is incorporated herein by reference in its entirety.

[0138] FIG. 2A shows a plurality of example designs of resonators 210, 220, 230, and 240. according to some embodiments. The resonators / arrays may be as described elsewhere herein. For example, the arrays may comprise one or more insulating materials. In another example, the arrays may comprise one or more nanogaps. Designs 220, 230, and 240 may comprise one or more nanogaps 201. The nanogaps may be as described elsewhere herein. For example, the nanogaps can be configured to concentrate anoptical field within the nanogap. For example, 210-240 of FIG. 2A depicts non-limiting examples of resonators with two or more nanostructures. In these depicted embodiments, the resonator comprises nanostructures such as nanostructure 204 and nanostructure 205, and the resonator is the entirety of each of 210, 220, 230, and 240 of FIG. 2A. As depicted in 210, a resonator may not comprise a nanogap 204. As depicted in 220, a nanostructure may comprise a nanogap 205 and a resonator may comprise a plurality of nanogaps. The resonator in this depicted embodiment comprises a nanogap in each nanostructure, as denoted by separation in the nanostructure (201). As depicted in 230, a resonator may comprise a plurality of nanostructures and a single nanogap. The nanostructures may be configured in a way as to concentrate an electromagnetic field. For example, the composition of the nanostructures may concentrate an electromagnetic field. As another example, the arrangement of the nanostructures may concentrate an electromagnetic field. As yet another example, the orientation of the nanostructures may concentrate an electromagnetic field. The nanostructure may concentrate the electromagnetic field to one or more nanogaps. In some embodiments, the nanostructure may comprise silicon. In some embodiments, the nanogap may comprise silicon. In some embodiments, one or more nanostructures comprise one or more metallic compositions. In some embodiments, one or more nanogaps comprise one or more metallic compositions. In some embodiments, one or more nanogaps and one or more nanostructures comprise one or more metallic compositions. The one or more metallic compositions may be on top of the nanostructure, as depicted in FIG. 38. In FIG. 38. silicon nanostructures are depicted with triangles indicating the deposition of metal. In this depicted embodiment, a resonator with a plurality of nanogaps may only have metal at one of the plurality of nanogaps. In some embodiments, metal may be deposited around more than one of the plurality of nanogaps in a resonator. The side view in FIG. 39 depicted the metal deposition atop the nanostructure may be deposited only on the nanostructure and surrounding the nanogap or it may be deposited to protrude over the nanogap. In some embodiments, the metal may go across the nanogap such as a bridge across and / or over the nanogap. In some embodiments, the metal is directly in contact with the semiconductor. In some embodiments, it is separated from the semiconductor by a thin dielectric layer. The one or more metallic compositions may be disposed in the nanogap. In FIG. 38, additional variations of the placement of metal are depicted. In this depicted embodiment, the metal may be disposed in the nanogap 3810. Hie metal may be disposed on a raised surface in the nanogap 3820. The metal may be disposed in such a way as to form a nanogap inside a nanogap 3850. The metal may be disposed in a nanogap that is filled with a combination of the metal and another dielectric. The metal may be embedded in the other dielectric 3830 or the metal may be embedded through the substrate 3840. FIG. 40 shows different configurations of the metal nanostructure with differing shapes and locations around and in the nanogap. The presence of the nanostructure(s) can provide local field enhancements beyond those generated by the nanogap alone. Forexample, the triangular nanostructures can further enhance the local field while maintaining the volume of the nanogap. The presence of the ovate nanostructure can provide even higher field enhancements as well as reduce the volume the field enhancements occur in, which may provide for single molecule signal generation as described elsewhere herein. The array can comprise one or more regions 202 configured as photonic mirror structures. The photonic mirror structures can be configured to couple far field light into the array. Hie array can comprise a cavity structure 203. The cavity structure can be configured to concentrate the field into the nanogap or slot as described elsewhere herein. FIG. 2B shows an example of an alternative view of design 220, according to some embodiments. As described elsewhere herein, the individual features of the arrays may comprise features with a height, width, and length independently selected from one another. The height, width, or length may be at least about 1 nanometer (nm), 5 nm, 10 nm, 25 nm, 50 nm, 75 nm, 150 nm, 200 nm 250 mn. 300 nm, 350 nm. 400 nm, 450 nm, 500 nm, 550 nm, 600 nm. 650 nm, 700 nm, 750 nm, 800 nm, 850 nm. 900 nm, 950 nm. 1 micrometer (pm), 2pm. 3pm. 4pm, 5pm, 6 pm, 7pm, 8pm, 9pm, 10 pm, 15 pm, 20 pm, 25 pm, 30 pm, 35 pm, 40 pm, 45 pm, 50 pm, 55 pm, 60 pm, 65 pm, 70 pm, 75 pm, 80 pm, 85 pm, 90 pm, 95 pm, 100 pm, 150 pm, 200 pm, 250 pm, 300 pm, 350 pm, 400 pm, 450 pm, 500 pm, 550 pm, 600 pm, 650 pm, 700 pm, 750 pm, 800 pm, 850 pm, 900 pm, 950 pm, or more. The height, width, or length may be at most about 950 micrometers (pm), 900 pm. 850 pm, 800 pm. 750 pm, 700 pm, 650 pm, 600 pm, 550 pm, 500 pm, 450 pm, 400 pm, 350 pm, 300 pm, 250 pm. 200 pm, 150 pm. 100 pm, 95 pm, 90 pm, 85 pm, 80 pm. 75 pm. 70 pm. 65 pm. 60 pm, 55 pm, 50 pm, 45 pm, 40 pm, 35 pm, 30 pm, 25 pm, 20 pm, 15 pm, 10 pm, 9 pm, 8pm, 7pm, 6pm, 5pm, 4pm, 3pm, 2pm, 1pm, 950 nanometers (nm), 900 mn, 850 nm, 800 nm, 750 nm, 700 nm, 650 nm, 600 nm, 550 nm, 500 mn, 450 nm, 400 nm, 350 nm, 300 nm, 250 nm, 200 nm, 150 nm, 100 nm, 75 nm, 50 nm, 25 nm, 10 nm, 1 nm, or less. The features may be separated by a distance of at least about 1 nanometer (mn), 5 nm, 10 nm, 25 nm. 50 nm, 75 nm, 150 nm. 200 nm 250 nm, 300 nm. 350 nm, 400 nm, 450 nm. 500 nm, 550 mn. 600 nm, 650 nm, 700 nm. 750 nm, 800 nm. 850 nm, 900 nm. 950 nm, 1 micrometer (pm), 2pm, 3pm, 4pm, 5pm, 6pm, 7pm, 8pm, 9pm, 10 pm, 15 pm, 20 pm, 25 pm, 30 pm, 35 pm, 40 pm, 45 pm, 50 pm, 55 pm, 60 pm, 65 pm, 70 pm, 75 pm, 80 pm, 85 pm, 90 pm, 95 pm, 100 pm, 150 pm, 200 pm, 250 pm, 300 pm, 350 pm, 400 pm, 450 pm, 500 pm, 550 pm, 600 pm, 650 pm, 700 pm, 750 pm, 800 pm. 850 pm, 900 pm, 950 pm, or more. The features may be separated by a distance (e.g., have a gap distance or slot size) of at most about 950 micrometers (pm), 900 pm. 850 pm, 800 pm. 750 pm, 700 pm. 650 pm, 600 pm. 550 pm, 500 pm, 450 pm. 400 pm, 350 pm. 300 pm, 250 pm, 200 pm, 150 pm, 100 pm, 95 pm, 90 pm, 85 pm, 80 pm, 75 pm, 70 pm, 65 pm, 60 pm, 55 pm, 50 pm, 45 pm, 40 pm, 35 pm, 30 pm, 25 pm, 20 pm, 15 p.m, 10 p.m, 9 p.m, 8 p.m, 7 p.m, 6 p.m, 5 p.m, 4pm, 3pm, 2pm, 1pm, 950 nanometers (nm), 900 nm, 850 nm, 800 nm, 750 nm, 700 nm, 650 nm, 600nm, 550 nm 500 nm, 450 nm, 400 nm, 350 nm, 300 nm, 250 nm, 200 nm, 150 nm, 100 nm, 75 nm, 50 nm, 25 nm, 10 nm, 1 nm, or less.

[0139] FIGS. 3 A - 3F show alternative designs for arrays of non-uniform features, according to some embodiments. Array 310 of FIG. 3 A shows a plurality of arrays of non-uniform features placed between features 311. The features 311 may be configured to decouple the modes of the arrays. For example, the individual sensing arrays can be configured to detect a biological molecule as described elsewhere herein, and the features can reduce or eliminate the crosstalk (e.g., reduce leakage of the light fields between the individual sensing arrays). The features 311 may comprise one or more dielectric materials. The features may comprise at least about 1. 2, 3, 4. 5. 6, 7, 8. 9, 10. or more features. The features may comprise at most about 10, 9, 8, 7, 6, 5, 4, 3, 2, or fewer features. The sensing arrays may be as described elsewhere herein. For example, the sensing arrays may comprise one or more nanogaps. The presence of the features 311 may enable close spacing of the sensing arrays. For example, the presence of the features can enable reduced cross talk in arrays spaced less than about 2 micrometers from one another. FIG. 3B shows a plurality of example designs of arrays 320, 330, 340, and 350 of non-unifonn features. Hie various designs provide different perturbations can provide different fields. The perturbations can provide benefits such as, for example, decoupling adjacent array’s fields, adjusting the field concentration, providing different coupling profiles, or the like. The perturbations may be perturbations or modulations of the dimensions, positions, angles, shapes, heights, or the like, of the features of the array.

[0140] FIG. 3C shows an example of an array of non-uniform features geometrically configured to reduce cross talk, according to some embodiments. The widths of the array of non-uniform features can be adjusted to decouple the modes carried by each of the arrays. For example, the width of a first array can be set such that the wavelength of light that the first array is configured to interact with is different from tire wavelength of light a second adjacent array is configured to interact with. In this example, the two arrays can be positioned within a wavelength of one another without interaction of the light between the two arrays. In this example, less than one micrometer spacings can be achieved without interference between the arrays. The arrays may comprise photonic cr stal mirrors. The photonic cry stal mirrors maybe configured to couple incident light into the resonant modes of the array. For example, the ends of a one-dimensional array may be configured as photonic crystal mirrors. The arrays may be periodic arrays. For example, the arrays can have a repeating structure (e.g.. a pattern to the dimensions of the elements of the arrays). The arrays may be aperiodic (e.g., without repeating structure). The periodicity or aperiodicity may be in tire width of the arrays. For example, an aperiodic array can have elements of the same height but different widths.

[0141] FIG. 3D shows an alternative set of isolating features 371 , according to some embodiments. The isolating features can serve a similar role as features 311 of FIG. 3A (e.g., isolating arrays from one another). FIGS. 3E and 3F show alternate periodic array structures, according to some embodiments. The different resonators of the arrays can be interleaved such as the resonators of FIG. 3E or spatially distinct such as the resonators of FIG. 3F.

[0142] FIG. 4 shows an example of a two-dimensional array of non-uniform features 400, according to some embodiments. Tire array may comprise one or more gaps 401. The gaps may be as described elsewhere herein. For example, the gaps can be configured to concentrate a light field within the gap. The two-dimensional array can be configured to provide a plurality of sensing regions in a small footprint by increasing the density of gaps that can be achieved. For example, the offset nature of the gaps of the array 400 can provide individual sensing regions configured for minimal interference while reducing spacings between the gaps. Any of the arrays of the present disclosure may be suitable for use in a two- dimensional array. For example, any of the arrays of FIGS. 2 - 3 can be configured as a two-dimensional array. The arrays may be separated by a distance of at least about 1 nanometer (nm), 5 nm, 10 nm, 25 nm, 50 nm, 75 nm, 150 nm, 200 nm 250 nm, 300 nm, 350 nm, 400 nm, 450 nm, 500 nm, 550 nm, 600 nm, 650 nm, 700 nm, 750 nm, 800 nm, 850 nm, 900 nm, 950 nm, 1 micrometer (pm), 2pm, 3pm, 4pm, 5pm, 6 vim, 7 vim, 8 vim, 9 vim, 10 vim, 15 vim, 20 vim, 25 pm, 30 pm, 35 pm, 40 pm, 45 pm, 50 pm, 55 pm, 60 pm, 65 pm, 70 pm, 75 pm, 80 pm, 85 pm, 90 pm, 95 pm, 100 pm, 150 pm, 200 pm, 250 pm, 300 pm, 350 pm, 400 pm. 450 pm, 500 pm, 550 pm, 600 pm, 650 pm, 700 pm, 750 pm, 800 pm, 850 pm, 900 pm, 950 pm, or more. Hie arrays may be separated by a distance of at most about 950 micrometers (pm), 900 pm, 850 pm, 800 pm, 750 pm, 700 pm, 650 pm, 600 pm, 550 pm, 500 pm, 450 pm, 400 pm, 350 pm, 300 pm, 250 pm, 200 pm, 150 pm, 100 pm, 95 pm, 90 pm, 85 pm, 80 pm, 75 pm, 70 pm, 65 pm, 60 pm, 55 pm, 50 pm, 45 pm, 40 pm, 35 pm, 30 pm, 25 pm, 20 pm, 15 pm, 10 pm, 9 pm, 8pm, 7pm, 6pm, 5pm, 4pm, 3pm, 2pm, 1pm, 950 nanometers (nm), 900 nm, 850 nm, 800 nm, 750 nm, 700 nm. 650 nm, 600 nm, 550 nm, 500 nm, 450 nm, 400 nm, 350 nm, 300 nm, 250 nm. 200 nm, 150 nm, 100 nm. 75 nm. 50 nm, 25 nm, 10 nm, 1 nm, or less.

[0143] FIGS. 5 A - 5B show examples of two-dimensional arrays 510 and 520, according to some embodiments. The two-dimensional arrays may comprise a plurality of features 511. The features may comprise shapes such as, for example, circles, polygons (e.g., triangles, squares, rectangles, trapezoids, diamonds, pentagons, etc.), lines, or the like, or any combination thereof. In the example of FIGS. 5 A - 5B, the features can be circles. The features can be configured to concentrate light fields as described elsewhere herein. For example, the array 510 can be an alternative to array 210 of FIG. 2A. As such, the arrays of the present disclosure may comprise any variation of the shape of the elements of the arrays.The array 520 can be configured with a gap 512. The gap may be as described elsewhere herein. For example, the gap can be functionalized with a capture probe for immobilizing a biological molecule within the gap. Tire gap may be a nanogap.

[0144] FIG. 6 A shows the portions of an array 600, according to some embodiments. The array may comprise one or more photonic mirror structures 601. The photonic mirror structures may be configured to couple light into the cavity structure 602. The array may comprise at least about 1, 2, 3, 4, 5, or more photonic mirror structures. Tire array may comprise at most about 5, 4, 3, 2, or 1 photonic mirror structures. The photonic mirror structures may be adjusted depending on the light the photonic mirror structure is configured to interact with. For example, the photonic mirror structure can be increased in size to interact with longer wavelengths of light. The cavity structure 602 may be configured to concentrate the light coupled by the photonic mirror structure as described elsewhere herein. For example, the cavity structure may comprise a gap. FIG. 6B is an example of the field enhancement of an array not comprising a nanogap, according to some embodiments. Tire field enhancement may be generated by a full field simulation of the modes of the structure. In the example of FIG. 6B, an electric field enhancement of 2,500 times may be observed, which can provide a Q of the cavity of about 80,000. The arrays of the present disclosure can provide electric field enhancements of at least about 10, 50, 100, 200, 300, 400, 500, 600, 700, 800, 900, 1,000, 1,100, 1,200, 1,300, 1,400, 1,500, 1,600, 1,700, 1,800, 1,900, 2,000, 2,100, 2,200, 2,300, 2,400, 2,500, 2,600, 2,700, 2,800, 2,900, 3,000, 3,500, 4,000, 4,500, 5,000, 6,000, 7,000, 8,000, 9,000, 10,000, or more times. The arrays of the present disclosure can provide electric field enhancements of at most about 10,000. 9,000, 8,000. 7,000, 6,000, 5,000, 4,500, 4.000, 3,500. 3,000, 2,900. 2.800, 2,700, 2.600, 2,500, 2.400, 2,300, 2.200. 2,100, 2,000. 1,900, 1,800. 1,700, 1,600, 1,500, 1,400, 1,300, 1,200, 1,100, 1,000, 900, 800, 700, 600, 500, 400, 300, 200, 100, 50, 10, or less times.

[0145] FIGS. 7A - 7B show an example of a field enhancement calculation 710 and an inset zoom 720 of an array 700 comprising aplurality of gaps 701, according to some embodiments. As described elsewhere herein, the plurality of gaps can be configured to concentrate a field of an incident light within the gaps. As can be seen in the calculations, the field can be largely confined to the gaps, thereby improving the signal that can be achieved. In the example of FIGS. 7A - 7B, the array can comprise a plurality of nanogaps. In some cases, each member of the array can comprise at least one nanogap. Each nanogap of the plurality of nanogaps can be functionalized as described elsewhere herein. For example, each nanogap can be configured with agents configured to bind one or more analytes. In some cases, an array can comprise a single nanogap. For example, FIG. 8A shows an example of an array 800 with a single nanogap 811, according to some embodiments . The inset 810 may show the nanogap region inincreased detail. The plot 820 of FIG. 8B is an example of an absolute field profile along the transverse direction of the array member comprising nanogap 811. As seen in 820, the high field confined to the center of the nanogap can provide the benefits described elsewhere herein (e.g., enhanced detection of an analyte configured within the nanogap).

[0146] FIG. 9 shows an example of a chip 910 comprising a detection region 920 and a separation region 930, according to some embodiments. The chip may comprise at least about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more detection regions. The chip may comprise at least about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more separation regions. The detection regions may comprise one or more arrays as described elsewhere herein. For example, the detection region may comprise at least one array of features comprising a nanogap. The separation region may be as described elsewhere herein. The separation region may comprise pillars of various sizes separated by various spacings configured to separate a sample by size. The pillars can comprise metals, dielectrics, insulators, or the like, or any combination thereof. The separation region may comprise one or more electrodes. Tire electrodes may be interdigitated (e.g., portions of the electrodes can be interspersed between other portions of the electrodes). Tire electrodes may be configured to separate analytes by charge. The pillars and / or substrate below the pillars may comprise one or more surface functionalizations and / or modifications. For example, the pillars can be coated with a nucleic acid sequence configured to bind and remove a non-target nucleic acid molecule from the sample.

[0147] FIGS. 10A - 10C show a pathway for analysis of the spectra of the present disclosure, according to some embodiments. Spectral data (e.g., Raman spectra, etc.), can be collected as described elsewhere herein and transferred to a database. The database may comprise additional information such as, for example, structure data, genetics data, sample data, or the like, or any combination thereof. The database may be used to train a machine learning algorithm. The machine learning algorithm can then be configured to analyze new data of an unknown sample to determine properties of the sample (e.g., presence or absence of analytes, structure of analytes, post-translational modifications, etc.). FIG. 10B shows an example of a structure mapping pathway, according to some embodiments. The structure mapping pathway can be configured, using one or more computer processors, to decompose a spectrum into one or more constituent signals. The constituent signals can correspond to structural motifs present in the analyte, which can provide information regarding the structural composition of the analyte. In the example of FIG. 10B. the different portions of the protein can generate Raman signals in different parts of the spectrum, which can then be used to identify the constituent portions of the analyte. In this way, recurring subunits or motifs can be identified in samples, which can provide information regarding analyte taxonomy and / or constellations. FIG. 10C shows an example of a structure prediction pathway,according to some embodiments. Using the spectra of the present disclosure, the structure of a new analyte can be predicted even if the structure is otherwise unknown. For example, based on the structure of a previously determined analyte, a new analyte can be analyzed and predictions for the component structures can be made. The kinetics and / or activity of the analyte can be predicted as well (e.g., the binding kinetics of a protein). In the example of FIG. IOC, a variety of moieties are being predicted for regions Bl - B4.

[0148] Tire arrays of the present disclosure may be configured to have controlled far field scattering. FIGS. 23A - 23C show far field scattering profiles of arrays, according to some embodiments. As seen in FIG. 23A. the far field scattering profile of an array shows that the array is highly isotropic in its scattering profile. Additionally, the scattering is concentrated on the array itself (the center of the radial plot), showing good localization to the array. In FIGS. 23B - 23C, the localization of the field to the arrays themselves can be seen. The separation of the fields can result in reduced crosstalk between the arrays and improved signal.Unit cells, photonic pillars, and multi-resonant chip designs

[0149] The chips of the present disclosure may be configured to comprise one or more unit cells. Said unit cells are patterned onto the chip in an array. This approach differs from longitudinally depositing parallel resonators onto a chip (e.g., a discus shape) and can improve efficiency of spectroscopic enhancement due to increased sensor packing density. A unit cell may comprise one or more resonators and / or one or more nanostructures. A resonator of a unit cell may be a photonic pillar and the terms are used interchangeably herein. A nanostructure of a unit cell may be referred to as an antenna and the terms used interchangeably herein. Photonic pillars function to enhance resonance in a unit cell on chip used in the methods as described herein (e.g., Section II).

[0150] The unit cell can have size dimensions (e.g., length and width) optimized based on analyte type, excitation and emission wavelengths, assay conditions, and / or purpose of the spectroscopic method for which the chip is designed. The length of a unit cell may be about 200 nm, about 250 nm, about 300 nm, about 350 nm, about 400 nm, about 450 nm, about 500 nm, about 550 nm, about 600 nm, about 650 nm, about 700 nm. about 750 nm, about 800 nm. about 850 nm, about 900 nm, about 950 nm, about 1000 nm, or as described in the Examples or Drawings. The width of a unit cell may be about 200 nm, about 250 nm, about 300 nm, about 350 nm, about 400 nm, about 450 nm, about 500 nm, about 550 nm, about 600 nm, about 650 nm, about 700 nm, about 750 nm, about 800 nm, about 850 nm, about 900 nm, about 950 nm, about 1000 nm, or as described in the Examples or Drawings. The length and width of the unit cell may be the same (e.g., a square unit cell shape) or different (e.g., a rectangular unit cell shape).

[0011] Unit cells comprise photonic pillars as resonators patterned in an array on the chip. A photonic pillar of a unit cell may have an elliptical or polygonal shape (e.g., when viewed top down). Photonic pillars may also have shapes with both linear and curved edges. In some embodiments, the elliptical shape of a photonic pillar is a circle. In some embodiments, the elliptical shape of a photonic pillar is an ellipse that is not a circle. In some embodiments, the shape of a photonic pillar is a polygon. Polygons have linear edges that can be easier to manufacture than curves. In some embodiments, the polygon is selected from a triangle, rectangle, pentagon, hexagon, heptagon, or octagon. A unit cell may comprise photonic pillars having a combination of shapes as described herein.

[0152] Photonic pillars can have size dimensions (e.g.. length, width, thickness) optimized based on analyte type, excitation and emission wavelengths, assay conditions, and / or purpose of the spectroscopic method for which the chip is designed. Tire length of a photonic pillar may be about 10 nm, about 20 mn, about 30 nm, about 40 nm, about 50 nm, about 60nm, about 70 mn, about 80 nm, about 90 nm, about lOOnm, about 110 nm, about 120 nm, about 130 nm, about 140 nm, about 150 nm, about 160 nm, about 170 nm. about 180 nm, about 190 nm. about 200 nm, about 210 nm. about 220, nm. about 230 nm, about 240 nm, about 250 nm, about 260 nm, about 270 nm, about 280 nm, about 290 nm, about 300 nm, about 350 nm, about 400 nm, or as described in the Examples or Drawings. The width of a photonic pillar may be 10 nm, about 20 nm, about 30 nm, about 40 nm, about 50 nm, about 60nm, about 70 nm, about 80 nm, about 90 nm, about lOOnm, about 110 nm, about 120 mn, about 130 nm, about 140 nm, about 150 nm, about 160 nm. about 170 nm, about 180 nm. about 190 nm, about 200 nm, about 210 nm, about 220, nm, about 230 nm. about 240 nm, about 250 nm. about 260 nm, about 270 nm, about 280 nm, about 290 nm, about 300 nm, about 350 nm, about 400 nm, or as described in the Examples or Drawings. The thickness of a photonic pillar may be about 10 nm, about 15 nm, about 20 nm, about 25 nm, about 30 nm, about 35nm, about 40 nm, about 45 nm, about 50 nm, about 60nm, about 70 nm, about 80 nm, about 90 nm, about lOOnm, about 110 nm, about 120 mn, about 130 nm, about 140 mn, about 150 nm, about 160 nm, about 170 nm. about 180 nm, about 190 nm. about 200 nm, about 250 nm, about 300 nm, about 350 nm, about 400 nm. about 450 nm, about 500 nm. about 550 nm, about 600 nm, about 650 nm, about 700 nm, about 750 nm, about 800 nm, about 850 nm, about 900 nm, about 950 nm, about 1000 nm, or as described in the Examples or Drawings.

[0153] A photonic pillar of a unit cell may be a primary photonic pillar. A primary photonic pillar is designed to have maximum resonance on the chip. A biological sample as described herein is generally placed at or near tire primary' photonic pillar. A photonic pillar may comprise one or more nanostructures. The nanostructures may increase spectroscopic resonance. The nanostructure of a primary' photonic pillar can be tuned for purpose (e.g., function as antennae). The nanostructures of a photonic pillar in a unit cellmay be an antennae, and the term is used interchangeably. A primary photonic pillar may have an optimized shape and size (e.g., length, width, radius, thickness). In some embodiments, the shape of the primary photonic pillar is about equal to the shape of a secondary photonic pillar. In some embodiments, the size of the primary photonic pillar is about equal to the size of a secondary photonic pillar in a chip design. In some embodiments, the shape and size of the primary photonic pillar is different from the secondary photonic pillar in the chip design.

[0154] Tire one or more photonic pillar of a unit cell may be a secondary photonic pillar. The secondaryphotonic pillar may improve resonance on the chip (e.g., at the primary photonic pillar or sample). The secondary photonic pillar may not have any nanostructures (e.g., antennae). In some embodiments, the shape of the primary photonic pillar is about equal to the shape of a secondary photonic pillar. In some embodiments, the size of the secondary photonic pillar is about equal to the size of a primary photonic pillar in a chip design. In some embodiments, the shape and size of the primary photonic pillar is different from the secondary photonic pillar in tire chip design.

[0015] A primary7photonic pillar of a unit cell may comprise on or more nanostructure (e.g., antennae). A primary7photonic pillar may comprise 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 nanostructures (e.g., antennae). Tire antennae may be orientated at the center of the primary photonic pillar, or off axis. Tire antennae may be rotated relative to a center line of the photonic pillar and array. A bowtie antenna may comprise two nanostructures. A rod antenna may consist of a single nanostructure Antennae of a primary photonic pillar may have a polygon or elliptical shape (e.g., when viewed top down). Antennae may also have shapes with both linear and curved edges. In some embodiments, the shape of an antennae is a polygon. Polygons have linear edges that can be easier to manufacture than curves. In some embodiments, the polygon of the antennae is selected from a triangle, rectangle, pentagon, hexagon, heptagon, or octagon. In some embodiments, the elliptical shape of an antennae is a circle. In some embodiments, the elliptical shape of an antennae is an ellipse that is not a circle. Antennae may comprise a combination of shapes as described herein. Each antennae may have the same or different shapes.

[0156] The shape and size of each nanostructure (e g., antenna) may be optimized. Antennae may comprise a metal, metal composition, or other highly resonance composition. The nanostructures (e.g., antennae) may have size dimensions (e.g., length, width, radius, curvature, thickness) optimized based on analyte type, excitation and emission wavelengths, assay conditions, and / or purpose of the spectroscopic method for which the chip is designed. Dimensions for antennae are also described herein as t(bowtie), t(rod), and t(thickness). Dimensions for antennae can vary based on the number of them present on the primary photonic pillar. The length or width (e.g., a bowtie length or rod length) of an antennae may beabout 10 nm, about 20 nm, about 30 nm, about 40 nm, about 50 nm, about 60nm, about 70 nm, about 80 nm, about 90 nm, about lOOnm, about 110 nm, about 120 nm, about 130 nm, about 140 nm, about 1 0 nm, about 160 nm, about 170 nm, about 180 nm, about 190 nm, about 200 nm, or as described in the Examples or Drawings. Bow ties are typically short than rods because multiple copies are fitting across the dimension of the photonic pillar. The thickness (e.g., t(thickness)) of an antennae may be about 1 nm, about 2 nm, about 3 nm. about 4 nm, about 5 nm, about 6 nm. about 7 nm, about 8 nm, about 9 nm, about 10 nm, about 15 nm, about 20 nm, about 25 nm, about 30 nm, about 35nm, about 40 nm, about 45 nm, about 50 nm, about 60nm, about 70 nm, about 80 nm, about 90 nm, about lOOnm, or as described in the Examples or Drawings. The curvature (e.g., r(curvature)) of an antennae may be about 0.1 nm, about 0.2 nm, about 0.3 mn, about 0.4 nm, about 0.5 nm, about 0.6 nm, about 0.7 nm, about 0.8 mn, about 0.9 nm, about 1 nm, about 1.5 nm, about 2.0 mn, about 5 nm, ab about 10 nm, about 15 nm, about 20 nm, about 25 nm, about 30 nm, about 35nm, about 40 nm, about 45 nm, about 50 nm, or as described in the Examples or Drawings. Each antennae may have the same or different dimensions. The antennae may have one or more of the dimensions as shown in TABLE 1.Chip designs

[0157] The chips as described herein (e.g., photonic chips) may be mult resonant designs, in which the total resonance is a combination of resonator and antennae elements. The chips may comprise dielectric materials (e.g., dielectrics for resonators; dielectric spacers; dielectric fill layers; etc.), substrates (optionally silicon or glass wafers), mirrors; oxide layers, and other components described herein the Examples and Drawings, in any combination thereof.

[0158] A chip as described herein (e.g., for use in a method as described herein) may comprise a substrate layer and an array of one or more photonic unit cells. Tire array of the one or more photonic unit cells may be positioned above and adjacent to the substrate layer, and each of the photonic unit cells may comprise a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap. The primary photonic pillar may be an ellipse or polygon comprising a dielectric base layer, dielectric spacer layer, and one or more photonic antennae. The dielectric base layer of the primary photonic pillar may be positioned above and adjacent to the substrate layer; the dielectric spacer layer of the primary photonic pillar may be positioned above and adjacent to the dielectric base layer of the primary photonic pillar; and the one or more photonic antennae may be positioned above and adjacent to the dielectric spacer layer of the primary photonic pillar. Each of the one or more secondary photonic pillars may be an ellipse or polygon comprising a dielectric base layer. Hie dielectric base layer of the one or more secondaryphotonic pillars may be positioned above and adjacent to the substrate layer. The chip may have an arrangement of components as shown in the schematic of FIG. 2B.

[0159] The ratio of primary photonic pillars to secondary photonic pillars in a unit cell as described herein may be 1: 1, 1:2, 1:3, 1:4, 1:5, 1:6, 1:7, 1:8, 1.9, or 1:2. The ratio of primary photonic pillars in a unit cell as described herein may be 2: 1, 3: 1, 4: 1, or 5: 1.

[0160] A chip as described herein (e.g.. for use in a method as described herein) may comprise a substrate layer and an array of one or more photonic unit cells, wherein the substrate layer comprises a silicon layer and an oxide layer. The array of the one or more photonic unit cells may be positioned above and adjacent to the substrate layer, and each of the photonic unit cells may comprise a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap. The primary photonic pillar may be an ellipse or polygon comprising a dielectric base layer, dielectric spacer layer, and one or more photonic antennae. The dielectric base layer of the primary photonic pillar may be positioned above and adjacent to the substrate layer; the dielectric spacer layer of tire primary photonic pillar may be positioned above and adjacent to the dielectric base layer of the primary photonic pillar; and the one or more photonic antennae may be positioned above and adjacent to the dielectric spacer layer of the primary photonic pillar. Each of the one or more secondary photonic pillars may be an ellipse or polygon comprising a dielectric base layer and dielectric spacer layer. The dielectric base layer of the one or more secondary photonic pillars may be positioned above and adjacent to the substrate layer; and the dielectric spacer layer of the one or more secondary photonic pillars may be positioned above and adjacent to the dielectric base layer of the one or more secondary’ photonic pillars. Tire substrate layer may comprise a silicon layer and an oxide layer, wherein the oxide layer is above and adjacent to tire silicon layer. Hie chip may have an arrangement of components as shown in the schematic of FIG. 68A.

[0161] A chip as described herein (e.g., for use in a method as described herein) may comprise a substrate layer and an array of one or more photonic unit cells, wherein the substrate layer comprises a metal layer and an oxide layer. The array of the one or more photonic unit cells may be positioned above and adjacent to the substrate layer, and each of the photonic unit cells may comprise a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap. The primary photonic pillar may be an ellipse or polygon comprising a dielectric base layer, dielectric spacer layer, and one or more photonic antennae. The dielectric base layer of the primary photonic pillar may be positioned above and adjacent to the substrate layer; the dielectric spacer layer of the primary photonic pillar may be positioned above andadjacent to the dielectric base layer of the primary photonic pillar; and the one or more photonic antennae are positioned above and adjacent to the dielectric spacer layer of the primary photonic pillar. Each of the one or more secondary photonic pillars may be an ellipse or polygon comprising a dielectric base layer and dielectric spacer layer. Tire dielectric base layer of the one or more secondary photonic pillars may be positioned above and adjacent to the substrate layer; and the dielectric spacer layer of the one or more secondary photonic pillars may be positioned above and adjacent to the dielectric base layer of the one or more secondary photonic pillars. The substrate layer may comprise a metal layer and an oxide layer, wherein the oxide layer is above and adjacent to the metal layer. The chip may have an arrangement of components as shown in the schematic of FIG. 68B.

[0162] A chip as described herein (e.g., for use in a method as described herein) may comprise a substrate layer, a dielectric fill layer, and an array of one or more photonic unit cells. Tire array of the one or more photonic unit cells may be positioned above and adjacent to the substrate layer, and each of the photonic unit cells comprises a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap. The primary photonic pillar may be an ellipse or polygon comprising a dielectric base layer, a portion of the dielectric fdl layer, and one or more photonic antennae. The dielectric base layer of the primary photonic pillar may be positioned above and adjacent to the substrate layer; the portion of the dielectric fill layer of the primary photonic pillar may be positioned above and adjacent to the dielectric base layer of the primary photonic pillar, and the one or more photonic antennae may be positioned above and adjacent to the portion of the dielectric fill layer of the primary photonic pillar. Each of the one or more secondary photonic pillars may be an ellipse or polygon comprising a dielectric base layer and a portion of the dielectric fill layer. The dielectric base layer of the one or more secondary photonic pillars may be positioned above and adjacent to the substrate layer; and the portion of the dielectric fill layer of the secondary photonic pillar may positioned above and adjacent to the dielectric base layer of the secondary photonic pillar. Tire dielectric fill layer may cover the surface of and the gap between each of the primary and secondary photonic pillars and may be below the one or more photonic antennae of the primary photonic pillar. The chip may have an arrangement of components as shown in the schematic of FIG. 68C.

[0163] A chip as described herein (e.g.. for use in a method as described herein) may comprise a substrate layer, a dielectric fill layer, an array of one or more photonic unit cells, and a passivation layer. The array of the one or more photonic unit cells may be positioned above and adjacent to the substrate layer, and each of the photonic unit cells may comprise a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondaryphotonic pillars are separated by a gap. Tire primary photonic pillar may be an ellipse or polygon comprising a dielectric base layer, a portion of the dielectric fill layer, one or more photonic antennae, and a portion of the passivation layer. The dielectric base layer of the primary photonic pillar may be positioned above and adjacent to the substrate layer; the portion of the dielectric fill layer of the primary photonic pillar may be positioned above and adjacent to the dielectric base layer of the primary photonic pillar; the one or more photonic antennae may be positioned above and adjacent to the portion of the dielectric fill layer of the primary photonic pillar; and the portion of the passivation layer of the primary photonic pillar may be positioned above and adjacent to the one or more photonic antennae and the portion of the fill layer of the primary photonic pillar. Each of the one or more secondary photonic pillars may be an ellipse or polygon comprising a dielectric base layer, a portion of the dielectric fill layer, and a portion of the passivation layer. The dielectric base layer of the one or more secondary photonic pillars may be positioned above and adjacent to the substrate layer; the portion of tire dielectric fill layer of the secondary photonic pillar may be positioned above and adjacent to the dielectric base layer of the secondary? photonic pillar; the portion of the passivation layer of the secondary photonic pillar may be positioned above and adjacent to the portion of the dielectric fill layer of the secondary' photonic pillar. The dielectric fill layer may cover the surface of and the gap between each of the primary and secondaryphotonic pillars and may be below the one or more photonic antennae of the primary photonic pillar, and the passivation layer may cover the surface of the dielectric fill layer. The passivation layer may comprise a gap in coverage at the one or more photonic antennae, wherein the gap in coverage is an absence of passivation layer. The chip may have an arrangement of components as shown in the schematic of FIG. 68D.

[0164] A chip as described herein (e.g., for use in a method as described herein) may comprise a substrate layer, a dielectric fill layer, an array of one or more photonic unit cells, wherein thesubstrate layer comprises a metal layer or dielectric mirror layer. The array of the one or more photonic unit cells may be positioned above and adjacent to the substrate layer, and each of the photonic unit cells may comprise a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap. The primaryphotonic pillar may be an ellipse or polygon comprising a dielectric base layer, a portion of the dielectric fill layer, and one or more photonic antennas. The dielectric base layer of the primary' photonic pillar may be positioned above and adjacent to the substrate layer; the portion of the dielectric fill layer of the primary photonic pillar may be positioned above and adjacent to the dielectric base layer of the primary photonic pillar, and the one or more photonic antennae may be positioned above and adjacent to the portion of the dielectric fill layer of the primary photonic pillar. Each of the one or more secondaryphotonic pillars may be an ellipse or polygon comprising a dielectric base layer and a portion of the dielectric fill layer. Tire dielectric base layer of the one or more secondary photonic pillars may be positioned above and adjacent to the substrate layer; the portion of the dielectric fill layer of the secondary photonic pillar may be positioned above and adjacent to the dielectric base layer of the secondary photonic pillar. The dielectric fill layer may cover the surface of and the gap between each of the primary and secondary photonic pillars and is below the one or more photonic antennae of the primary photonic pillar. The substrate layer may comprise a metal layer or dielectric mirror layer and dielectric spacer layer. The dielectric spacer layer may be above and adjacent to the metal layer or dielectric mirror layer. The chip may have an arrangement of components as shown in the schematic of FIG. 69A.

[0165] A chip as described herein (e.g., for use in a method as described herein) may comprise a substrate layer, a dielectric fill layer, an array of one or more photonic unit cells, wherein the substrate layer comprises a silicon or glass wafer layer, a metal layer or dielectric mirror layer, and a dielectric spacer layer a mirror or mirror-like surface. The array of the one or more photonic unit cells may be positioned above and adjacent to the substrate layer, each of the photonic unit cells may comprise a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap. The primary photonic pillar may be an ellipse or polygon comprising a dielectric base layer, a portion of the dielectric fdl layer, and one or more photonic antennae. The dielectric base layer of the primary photonic pillar may be positioned above and adjacent to the substrate layer; the portion of the dielectric fill layer of the primary photonic pillar may be positioned above and adjacent to the dielectric base layer of the primary photonic pillar, and the one or more photonic antennae may be positioned above and adjacent to the portion of the dielectric fill layer of the primary photonic pillar. Each of the one or more secondary photonic pillars may be an ellipse or polygon comprising a dielectric base layer and a portion of the dielectric fill layer. The dielectric base layer of the one or more secondary photonic pillars may be positioned above and adjacent to the substrate layer; and the portion of the dielectric fill layer of the secondary photonic pillar may be positioned above and adjacent to the dielectric base layer of the secondary photonic pillar. The dielectric fill layer mayt covthe surface of and the gap between each of the primary and secondaryphotonic pillars and is below the one or more photonic antennae of the primary photonic pillar, and the substrate layer comprises a silicon or glass wafer layer, a metal layer or dielectric mirror layer, and a dielectric spacer layer, wherein the metal layer or dielectric mirror layer is above and adjacent to the silicon or glass wafer layer, and the dielectric spacer layer is above and adjacent to the metal layer or dielectric mirror layer. The chip may have an arrangement of components as shown in the schematic of FIG. 69D.

[0166] chip as described herein (e.g., for use in a method as described herein) may comprise a substrate layer, a dielectric fdl layer, and an array of one or more photonic unit cells. The array of the one or more photonic unit cells may be positioned above and adjacent to the substrate layer, each of the photonic unit cells may comprise a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap. The primary photonic pillar may be an ellipse or polygon comprising a dielectric base layer, a portion of the dielectric fdl layer, and one or more photonic antennae. The dielectric base layer of the primary’ photonic pillar may be positioned above and adjacent to the substrate layer; the portion of the dielectric fdl layer of the primary’ photonic pillar may be positioned above and adjacent to the dielectric base layer of the primary photonic pillar, and the one or more photonic antennae may be positioned above and adjacent to the portion of the dielectric fill layer of the primary photonic pillar. Each of the one or more secondary photonic pillars may be an ellipse or polygon comprising a dielectric base layer and a portion of the dielectric fill layer. Tire dielectric base layer of the one or more secondary photonic pillars may be positioned above and adjacent to the substrate layer; and the portion of the dielectric fdl layer of the secondary photonic pillar may be positioned above and adjacent to the dielectric base layer of the secondary photonic pillar. The dielectric fill layer may cover the surface of and the gap between each of the primary and secondary photonic pillars and may be below the one or more photonic antennae of the primary photonic pillar. The substrate layer may comprise a silicon or glass wafer layer, alternating layers of a silicon layer and an oxide layer, and a dielectric spacer layer. The alternating layers of a silicon layer and an oxide layer may be above and adjacent to the silicon or glass wafer layer, and the dielectric spacer layer may be above and adjacent to tire alternating layers of a silicon layer and an oxide layer. Tire chip may have an arrangement of components as shown in the schematic of FIG. 69D.

[0167] Tire chip may be a single resonance chip, or a multi-resonance chip (e.g., two, three, or more). Resonance may come from one or more resonator and / or antennae designs on the same chip. Tire resonances may be synergistic.

[0168] Dielectric materials for chips may be selected from (but not limited to) dielectric materials such as pure silicon, crystalline silicon, amorphous silicon, silicon nitride, silicon nitride, silicon dioxide, silicon carbide, or fused silica. Materials for nanostructures (e.g., antennae) for chips may be selected from (but not limited to) metals and / or metal compositions selected from gold, silver, platinum, copper, aluminum, titanium nitride, gallium nitride, germanium, chromium. Materials for oxides may be selected from hafnium oxide, titanium oxide, ethylene oxide, nickel oxide, copper oxide, zinc oxide, indium oxide, gallium oxide, and optionally silicon oxides.Manufacturing stepsDual resonant chip designs as described herein are produced using semiconductor manufacturing methodologies (e.g., FIG. 70A-70Q). A method of manufacturing a chip as described herein may comprise one or more (e.g., up to all) of the steps of: (a) providing a starting substrate; (b) depositing a mirror or mirror-like layer on top of the starting substrate, wherein the mirror layer comprises one or more alternating layers of a silicon layer and an oxide layer, or a mirror layer; (c) depositing a dielectric spacer on top of the mirror layer; (d) depositing a dielectric device layer on top of the dielectric spacer layer, optionally wherein the dielectric device layer is a silicon device layer; (e) depositing a first mask layer on top of the dielectric device layer; (f) depositing a first photoresist layer on top of the first mask layer; (g) performing photolithography patterning on the first photoresist layer; (h) developing the first photoresist layer; (i) etching the first mask layer and the dielectric device layer; (j) removing the first mask; (k) depositing a dielectric fill layer on top of the exposed dielectric spacer and on the exposed dielectric device layer; (1) depositing layer of material for one or more photonic antennae on top of the dielectric fill layer; (m) depositing a second mask layer on top of the layer of material for the one or more photonic antennae; (n) depositing a second photoresist layer on top of the second mask layer; (o) performing photolithography patterning on the second photoresist layer; (p) developing the second photoresist layer; (q) etching the second mask layer and the layer of material for one or more photonic antennae; or (r) removing the second mask layer. Said method thereby produces a chip for use in a method as described herein. The chip may be manufactured as shown herein the Examples and Drawings.(b) Biological samples

[0169] Tire systems described herein may be used for the analysis of a biological sample. In An analyte in a method as described herein may be biological sample. In some embodiments, the sample is a liquid. In some embodiments, the sample is a dissolved solid. Though described herein with regards to biological samples, various other types of samples can be utilized with the methods and systems of the present disclosure. For example, samples from reactors can be utilized (e.g., samples used in line or batch from catalytic reactors (e.g., to generate polymers or other chemicals)). In other cases, environmental samples (e.g., water, soil, air, etc.) can be used. In other cases, food samples can be analyzed for, for example, a presence or absence of an adulterant, presence or absence of a key analyte, etc. Similarly, samples can be processed to identify low-concentration portions of the sample (e.g.. in forensic samples, etc.).

[0170] Tire biological sample may be a tissue sample. Tire biological sample may be a single cell (e.g., an immune cell). The biological sample may be a whole cell. The cell may be a mammalian cell.preferably a human cell. The cell may be an immune cell. In some embodiments, the immune cell is selected from a T cell, B cell, natural killer (NK) cell, neutrophil, eosinophil, basophil, mast cell, monocyte, or dendritic cell.

[0171] . The biological sample may be a plurality of cells. The biological sample may be biological components eluted from a cell. Non-limiting examples of biological components include, but are not limited to proteins, peptides, and nucleic acids. The peptide may be an antigen. The biological sample may also be a protein complex (e.g., an MHC-antigen complex). Non-limiting examples of “sample” include any material from which nucleic acids and / or proteins can be obtained. As non-limiting examples, this includes whole blood, peripheral blood, plasma, serum, saliva, mucus, urine, semen, lymph, fecal extract, cheek swab, cells or other bodily fluid or tissue, including but not limited to tissue obtained through surgical biopsy or surgical resection. The biological sample may comprise an organism, including without limitations, a bacterium or a virus. Tire biological sample may comprise a cell fragment. The cells may be eukaryotic. The cells may be prokaryotic. Tire biological sample may be a biological molecule. The biological sample may be a single biological molecule.

[0172] In some cases, the biological sample is a tissue, tissue homogenate, or organoid sample. In some cases, the biological sample is a collection of cells, or a single cell, or cell fragment. In some cases, tire sample comprises polynucleotides, such as DNA or RNA. In some cases, the sample comprises macromolecules, such as proteins, including antibodies. In some cases, the sample comprises polypeptides or peptides. In some cases, the sample comprises metabolites or other small molecules. In some cases, the sample may comprise one or more viruses. In some cases, the sample may comprise one or more micro-organisms, like bacteria. In some cases, the sample may comprise mixtures or conjugates of one or more of the above example samples (e.g., antibody -drug conjugates, DNA-barcoded proteins, etc). In some cases, the sample may be a mixture of molecules, or polymers, or microplastics.

[0173] In some embodiments, the sample comprises at least one component. In some embodiments, the sample comprises two or more components. In some embodiments, tire sample comprises, at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15. 16, 17, 18, 19, 20 or more components.

[0174] Tire component may be a nucleic acid. The nucleic acid may be DNA, RNA, or a combination thereof. The nucleic acid may be an oligonucleotide. The component may be a peptide, a polypeptide, or a protein. Hie protein may be whole. The protein may be fragmented. Hie protein may be an antibody. The antibody may be whole. The antibody may be fragmented. The component may be a metabolite. The component may be a polymer. The component may comprise a bead (e.g., a microbead). Thecomponent may comprise a nanoparticle. The component may be a microplastic. The component may be a small molecule. The component may be a sugar. The sugar may be a monosaccharide. The sugar may be a disaccharide. The sugar may be a polysaccharide. Tire polysaccharide may be a glycan. Though described herein with regards to biological samples, other samples can be used in the methods and systems of the present disclosure. For example, non-biological samples can be analyzed for various non- biological analytes. Examples of non-biological samples include, but are not limited to. polymers, industrial chemicals, environmental samples, or the like.II. Methods and Molecules for Label-free Spectroscopy

[0175] In certain aspects, described herein are methods of using a chip as described herein. In some embodiments, the methods described herein comprise a method of processing a biological sample. Any of the methods as described herein may be used on analyte that is a protein, peptide, nucleic acid (e.g., DNA or RNA), or a cell. The cell may be a whole cell. The cell may be a mammalian cell, preferably a human cell. The cell may be an immune cell. In some embodiments, the immune cell is selected from a T cell, B cell, natural killer (NK) cell, neutrophil, eosinophil, basophil, mast cell, monocyte, or dendritic cell.

[0176] In some embodiments, the methods described herein comprise a method of detecting or identifying an analyte's interaction with a sample. In some embodiments, the methods described herein comprise a method of filtering a sample. In some embodiments, the methods described herein comprise a method of detecting or identifying an analyte in a sample. In some embodiments, the methods described herein comprise a method of building a library of detected analytes. In some embodiments, the methods described herein comprise a method of building a library of identified analytes. In some embodiments, the methods described herein comprise a method of storing a reference profile of an analyte.

[0177] In certain aspects, described herein is a method of detecting or identifying an analyte’s interactions with a sample. In some embodiments, the method comprises providing the analyte on a chip described herein. Tire chip may comprise an array of features, wherein a feature of said array of features comprises an electrical insulator or a semiconductor, wherein said feature comprises a nanogap. The nanogap may comprise a nanostructure. The nanostructure may be functionalized. The nanostructure may be metal. The nanostructure may be functionalized and metal. The features may be non-uniform.

[0178] In certain aspects, described herein is a method of filtering a sample as described herein. In some embodiments, the method comprises providing a biological sample on a chip as described herein. Hie chip may comprise an array of features configured to filter the one or more components according tocharge as described herein. The features may be non-uniform. The chip may be interspersed with a plurality of electrodes or functionalized features configured to filter the one or more components according to charge or size. The methods may further comprise using the chip to filter the sample.

[0179] In certain aspects, described herein is a method of detecting or identifying an analytes interaction with a sample. The methods may comprise providing the analyte on a chip as described herein. The methods may comprise introducing a sample on the chip, wherein the chip comprises the analyte. The methods may then comprise following the real-time interactions of the analyte with the sample. The sample may be a sample as described herein. In certain aspects, described herein is a method of detecting or identifying a molecule’s interactions with a sample. The methods may comprise introducing a sample on the chip, and then introducing chemicals to the surroundings. The methods may then comprise following the real-time interactions of the molecules in the surrounding media with the sample. The sample may be a sample as described herein. In certain aspects, described herein is a method of detecting or identifying a molecule’s interactions with a sample. Tire methods may comprise introducing a sample on the chip, and then introducing chemicals to the surroundings. The methods may then comprise following the real-time interactions of the molecules in the surrounding media with the sample. The sample may be a sample as described herein. The sample may comprise a protein. The sample may comprise a peptide. The sample may comprise a nucleic acid. The surrounding media may comprise chemicals designed to cleave the peptide, protein, or nucleic acid, such as Edman reagents or enzymes (peptidases). Tire sample may comprise a protein. The sample may comprise a small molecule. Tire analyte may comprise a protein. The analyte may comprise an enzyme. The analyte may comprise a kinase. The analyte may comprise a receptor. The analyte may comprise a tyrosine kinase. The analyte may comprise a Janus kinase 3. The analyte may comprise an epidermal growth factor receptor (EGFR). A protein may be intact. A protein may be fragmented. A protein may be analyzed for an isoform. The analyte may comprise a major histocompatibility complex (MHC) peptide. Tire MHC peptide may comprise an MHC class I peptide. Tire MHC peptide may comprise an MHC class II peptide. The analyte may comprise an antibody. The antibody may comprise a T-cell. An antibody may comprise a B- cell. An antibody may be intact. An antibody may be fragmented. An antibody may be analyzed for identification of post-translational modifications. The analyte may comprise a biologic. A biologic may be analyzed for post-translational modifications. A biologic may be analyzed for secondary structure. A biologic may be analyzed for tertiary' structure. The analyte may comprise nucleic acids. The nucleic acids may comprise RNA. The nucleic acids may comprise DNA. The nucleic acids may comprise complementary DNA (cDNA).

[0180] In some embodiments, the methods comprise introducing a sample as described herein on the chip. In some embodiments, the methods comprise introducing a sample as described herein on a resonator. In some embodiments, the method comprises bringing a sample in contact with a resonator. In some embodiments, a sample is drop cast on a chip. In some embodiments, a spectrum is collected of an analyte in a sample still in solution. In some embodiments, the sample drop cast on a chip is dried before collecting a spectrum. In some embodiments, drop casting a sample on a chip brings it in contact with a resonator, a pixel, or an antenna structure. In some embodiments, the method comprises exposing the chip to a first light. Exposing the chip to a first light may expose the sample to a first light. In some embodiments, the methods comprise, exposing the chip to a first light from a light source, such that the first light interacts with an array of non-unifonn features and is further concentrated in said nanogap. In some embodiments, the methods comprise detecting a second light from said array of non-uniform features subsequent to said array of non-uniform features being exposed to the first light. In some embodiments, the chip may be exposed to two light sources. This combination of light sources interacts with the array of non-unifonn features and are concentrated in said nanogap. In this embodiment, the method comprises detecting a second light different from the first two light sources from said array of non-unifonn features. In some embodiments, the methods comprise collecting a time series of the second light. In some embodiments, the second light yields infrared scattering signature associated with the analyte. In some embodiments, the second light yields a vibrational scattering signature associated with the analyte. In some embodiments, the vibrational scattering signature is a Raman spectrum. In some embodiments, the second light yields autofluorescence associated with the analyte. In some embodiments, the second light is autofluorescence associated with the analyte. In some embodiments, the chip is exposed to a third light. Exposing the chip to a third light may expose the sample to a third light.

[0181] In some embodiments, a light source of the first light is integrated with the chip. In some embodiments, a light source of the first light is not integrated with the chip. In some embodiments, a light source of the second light is integrated with the chip. In some embodiments, a light source of the second light is not integrated with the chip. Tire incident light may be an incident laser, a light emitting diode (LED) light, a lamp, or a combination thereof The incident light may be an LED. The incident light may be a lamp. The incident light may be a laser as described herein.

[0182] In some embodiments, the methods further comprise providing a detector. The detector may comprise a detector plane. The methods may comprise using the detector to scan wavelengths in the detector plane. In some embodiments, the second light is detected with an integrated spectrometer or filter system on the chip. In some embodiments, the second light is detected with an integrated spectrometer or filter system that is not integrated with the chip.

[0183] In some embodiments, the methods further comprise imaging at least a portion ofthe chip. In some embodiments, the imaging comprises super resolution imaging. Super resolution imaging includes, without limitations, structured illumination microscopy (SIM), entropy based super resolution imaging (ESI), stochastic optical reconstruction microscopy (STORM), super resolution optical fluctuation imaging (SOFI). and stimulated emission depletion microscopy (STED). In some embodiments, the super resolution imaging comprises SIM. In some embodiments, the super resolution imaging comprises ESI. In some embodiments, the super resolution imaging comprises STORM. In some embodiments, the super resolution imaging comprises SOFI. In some embodiments, the super resolution imaging comprises STED.

[0184] In some embodiments, the methods described herein further comprise producing one or more hyperspectral images. In some embodiments, each of the one or more hyperspectral images represents a distinct Raman signature. In some embodiments, the methods described herein further comprise collecting data from the one or more hyperspectral images.In some embodiments, the method further comprises scanning a biological sample to produce the hyperspectral image. The scanning may comprise a spatial scanning, spectral scanning, non-scanning, spatiospectral scanning, or any combination thereof.

[0185] Various methods of detection can be utilized with the methods and systems of the present disclosure. Examples of detection schemes include, but are not limited to, spontaneous Raman spectroscopy, Stimulated Raman spectroscopy, coherent anti-Stokes Raman spectroscopy, hyperspectral mapping using spectral fdters on the detection side of the sample, hyperspectral mapping using a fixed pump laser and a variable probe wavelength, super resolution Raman imaging (e.g., using, for example, structured illumination microscopy, entropy based super resolution imaging, stochastic optical reconstruction microscopy, super resolution optical fluctuation imaging, etc ), or the like.

[0186] In some cases, the methods and systems of the present disclosure can identify an analyte with an accuracy, sensitivity, or specificity of at least about 60, 65, 70, 75, 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 99.9. or more percent. In some cases, the accuracy, specificity, or sensitivity can be achieved without use of a label (e.g., in a label-free manner).

[0187] In some embodiments, the methods described herein further comprise developing a machine learning model. In some embodiments, the machine learning model is a neural network. In some embodiments, the neural network is a convolutional neural network (CNN). In some cases, the neural netw ork is a large language model (LLM). Machine learning algorithms implemented on a computer system as described elsewhere herein (e.g, a remote server) can synthesize the spectra generated by themethods and systems of the present disclosure. For example, a machine learning algorithm can be configured to synthesize a spectrum that closely approximates an empirical spectrum. The synthetic spectrum can be used to augment datasets and / or input data to further train the machine learning model. Tire machine learning model may determine an identity of a biomolecule (e.g., a protein, peptide, or nucleic acid). A machine learning algorithm can be trained to determine one or more modifications associated with a peptide, polypeptide, protein, or nucleic acid. As a non-limiting example, a machine learning algorithm or computer vision algorithm can determine the location and / or identity of the modification. A machine learning model may be trained to identify properties of a biomolecule. Nonlimiting examples of properties of a biomolecule include but are not limited to amino acid sequence, nucleic acid sequence, modifications, and secondary and higher order structures. Tire spectrum for use by the machine learning model may be generated through chemical, electrical, or physical perturbations. Non-limiting examples of perturbations include heating, charge or electric field perturbations. pH perturbations, or chemical binding and cleavages. The machine learning algorithm may estimate a three- dimensional structure. The machine learning algorithm may comprise input data to assist in the generation of a three-dimensional structure. Non-limiting examples of machine learning data include chemical bond energy and modifications. A machine learning algorithm can be configured to determine a sequence of a protein polypeptide, or nucleic acid in an absence of degrading the protein polypeptide, or nucleic acid. Still other machine learning algorithms can be configured to detennine a sequence of a molecule without labeling the molecule. The machine learning model may predict the number of molecules bound to the system disclosed herein. The prediction of molecules bound to the systems disclosed herein may include values for the number of molecules bound to the systems disclosed herein. The prediction of molecules bound to the systems disclosed herein may include probabilities of values for the molecules bound to the systems disclosed herein. Examples of machine learning models include, but are not limited to, a neural network (NN), convolutional NN, autoencoder, transformers, attention networks, graph networks, autoregressive networks, sequential networks, or any combination thereof

[0188] The machine learning algorithms can be supervised, semi-supervised, or unsupervised. A supervised machine learning algorithm can be trained using labeled training inputs, e.g.. training inputs with known outputs. The training inputs can be provided to an untrained or partially trained version of the machine learning algorithm to generate a predicted output. The predicted output can be compared to the known output, and if there is a difference, the parameters of the machine learning algorithm can be updated. A semi-supervised machine learning algorithm can be trained using a large number of unlabeled training inputs and a small number of labeled training inputs. An unsupervised machine learning algorithm, e.g., a clustering algorithm, can find previously unknown patterns in data sets without pre-existing labels. Other examples of machine learning algorithms that can be used to process the spectra generated by the methods and systems of the present disclosure are regression algorithms, decision trees, support vector machines, Bayesian networks, clustering algorithms, reinforcement learning algorithms, and the like.

[0189] An example of an application of a machine learning algorithm can be the determination of a sequence of a polypeptide. The machine learning algorithm can be trained to identify a single residue of the polypeptide based on consecutive spectra taken during a degradation operation. The machine learning algorithm can be trained to identify a sequence of a peptide, protein, or polypeptide. The machine learning algorithm can be trained to identify a post-translational modification of a peptide, protein, or polypeptide. The machine learning algorithm can be trained to identify a fold or structures of a peptide, protein, or polypeptide. The machine learning algorithm can be trained to identify, e.g., a stretch energy, strain energy, twist energy, or any combination thereof from a chemical bond of a peptide, protein, or polypeptide. For example, the machine learning algorithm can be trained with spectra taken during the removal of a known residue from a polypeptide, and the machine learning algorithm can subsequently identify when a similar residue is removed from a different polypeptide. In some cases, the machine learning algorithm can comprise an autoregressive or autoregressive-like model (e.g., where the previously predicted residue can provide context for the prediction of the current residue). For example, the autoregressive model can be trained to note patterns that can appear in polypeptide sequences and can favor residue identifications that fit within those patterns. In some cases, the machine learning algorithm can comprise a graph-like model (e.g.. a Markov random fields model). Such a model can predict an entire sequence of a polypeptide at once, using a single spectrum or a combination of spectra of a single molecule. For example, a number of spectra can be acquired, averaged, and the graph-like model can determine a sequence of the molecule using the averaged spectra. In this example, the determination of the sequence may be significantly faster than techniques that use labeling or degradation operations.

[0190] In certain aspects, described herein is a method of detecting or identifying an analyte in a biological sample. In some embodiments, the method comprises providing said biological sample on a chip comprising an array of non-uniform features, wherein a feature of said array of non-uniform features comprises an electrical insulator or a semiconductor, wherein said feature comprises a nanogap. In some embodiments, the method comprises exposing said chip to a first light from a light source, such that said first light interacts with said array of non-unifonn features and is further concentrated in said nanogap. In some embodiments, the method comprises detecting a second light from said array of non-uniform features subsequent to said array of non-uniform features being exposed to said first light. In some embodiments, the method comprises using said second light to detect or identify said analyte.

[0191] In certain aspects, described herein is a method of filtering a sample. In some embodiments, the method comprises providing a biological sample comprising one or more components on a chip, said chip comprising an array of non-uniform features configured to filter said one or more components according to size. In some embodiments, the method comprises, wherein said array of non-uniform features are interspersed with a plurality of electrodes or functionalized features configured to filter said one or more components according to charge or size; In some embodiments, the method comprises using said chip to filter said sample comprising one or more components.

[0192] In certain aspects, described herein is a method of detecting or identifying an analyte’s interactions with a sample. In certain embodiments, the method comprises providing said analyte on a chip comprising an array of non-uniform features, wherein a feature of said array of non-uniform features comprises an electrical insulator or a semiconductor, wherein said feature comprises a nanogap. In certain embodiments, the method comprises introducing a sample on said chip. In certain embodiments, the method comprises exposing said chip to a first light from a light source, such that said first light interacts with said array of non-unifonn features and is further concentrated in said nanogap. In certain embodiments, the method comprises detecting a second light from said array of non-uniform features subsequent to said array of non-uniform features being exposed to said first light. In certain embodiments, the method comprises collecting a time series of the second light.

[0193] The present disclosure provides a method of detecting or identifying an analyte in a biological or chemical sample, comprising providing said biological or chemical sample on a chip comprising two or more resonators, wherein each of the two or more resonators supports one or more guided modes; wherein each of the two or more resonators has a corresponding longitudinal perturbation, where at least one guided mode resonance is supported in each resonator; wherein an incident light is coupled to two or more of the guided mode resonances by the longitudinal perturbations of the resonators; a feature of said array of non-uniform features wherein each resonator comprises an electrical insulator or a semiconductor; wherein each resonator comprises at least one, wherein said feature comprises a nanogap configured to concentrate an incident light; wherein one or more regions of high electromagnetic field intensity are localized within and in proximity to each nanogap, whereby environmental sensing is provided. In some embodiments, the method comprises exposing said chip to a first light from a light source, such that said first light interacts with resonators and is further concentrated in said nanogaps. In some embodiments, the method comprises detecting a second light from resonators subsequent to said array of non-uniform features being exposed to said first light. In some embodiments, the method comprises using said second light to detect or identify said analyte.

[0194] In certain aspects, described herein is a chip-based method of filtering a sample prior to detecting or identifying a analyte and / or interactions between an analyte and a binding molecule, comprising a biological or chemical sample comprising one or more components on a chip, said chip comprising an array of non-uniform features configured to filter said one or more components according to size, and wherein said array of non-uniform features are interspersed with a plurality of electrodes or functionalized features configured to filter said one or more components according to charge, or size, or chemical / biological affinity. In some cases, the filtering is on chip filtering (e.g., filtering using one or more elements of a chip). In some embodiments, the filtering is off chip filtering (e.g., filtering the analytes before the analytes are introduced to the chip). Examples of off-chip filtering include, but are not limited to, microfiltration, ultrafiltration, nanofiltiation, reverse osmosis, size-exclusion chromatography, ion-exchange chromatography, affinity chromatography, liquid chromatography, high- performance liquid chromatography (HPLC), gas chromatography , paper chromatography, thin-layer chromatography, or the like, or any combination thereof. In some embodiments, the method comprises using said chip to filter said sample comprising one or more components. In some embodiments, the method comprises providing said filtered sample on a sensor region on the same chip, said sensor region comprising two or more resonators, wherein each of the two or more resonators supports one or more guided modes; wherein each of the two or more resonators has a corresponding longitudinal perturbation, where at least one guided mode resonance is supported in each resonator; wherein an incident light is coupled to two or more of the guided mode resonances by the longitudinal perturbations of the resonators; a feature of said array of non-uniform features wherein each resonator comprises an electrical insulator or a semiconductor; wherein each resonator comprises at least one, wherein said feature comprises a nanogap configured to concentrate an incident light; wherein one or more regions of high electromagnetic field intensity are localized within and in proximity to each nanogap, whereby environmental sensing is provided. In some embodiments, the method comprises exposing said sensor region to a first light from a light source, such that said first light interacts with resonators and is further concentrated in said nanogaps. In some embodiments, the method comprises detecting a second light from resonators subsequent to said array of non-uniform features being exposed to said first light. In some embodiments, the method comprises using said second light to detect or identify said analyte.

[0195] In certain aspects, described herein is a method of detecting or identifying an analyte’s interactions with a sample. In some embodiments, the method comprises providing said analyte on a chip comprising two or more resonators, wherein each of the two or more resonators supports one or more guided modes; wherein each of the two or more resonators has a corresponding longitudinal perturbation, where at least one guided mode resonance is supported in each resonator; wherein an incident light iscoupled to two or more of the guided mode resonances by the longitudinal perturbations of the resonators; a feature of said array of non-uniform features wherein each resonator comprises an electrical insulator or a semiconductor; wherein each resonator comprises at least one, wherein said feature comprises a nanogap configured to concentrate an incident light; wherein one or more regions of high electromagnetic field intensity are localized within and in proximity to each nanogap, whereby environmental sensing is provided. In some embodiments, the method comprises introducing a sample on said chip. In some embodiments, the method comprises exposing said chip to a first light from a light source, such that said first light interacts with said array of resonators. In some embodiments, the method comprises detecting a second light from resonators subsequent to said resonators being exposed to said first light. In some embodiments, the method comprises collecting a time series of the second light.

[0196]

[0197] FIG. 11 shows an example of tissue mapping with an array, according to some embodiments. Light 1101 may be light as described elsewhere herein. The light can be directed from a light source (not pictured) towards an array 1102 (e.g., an array as described elsewhere herein). The array can be configured to enhance the light field within the array, which can improve the signal that the array produces. A sample 1103 can be placed adjacent to the array. For example, an unstained fixed tissue section can be placed atop an array. A plurality of samples can be placed atop an array. The sample can be placed such that portions of the sample interact with the light fields concentrated by the array. For example, the sample can be placed such that a portion of the sample is in sufficient proximity to a gap in a member of the array as to provide enhanced light fields encompassing the portion of the sample. In some cases, the array can comprise a binding moiety configured to bind to an analyte within the tissue. For example, a gap within a member of the array can comprise a nucleic acid probe configured to bind to at least a portion of an analyte comprised within the sample. In this example, the analyte can move out of the sample and be bound to the probe. The light 1101 interacting with the array 1102 and the sample 1 103 can generate a hyperspectral map 1 104 comprise a plurality of spectra (e.g., spectra 1104 and 1 105). The hyperspectral map may provide information regarding the distribution of analytes or other features within the sample. For example, the spectra can provide identification for various analytes or structural features within the sample. In another example, the spectra can provide distribution data for analytes within the sample. The sample (e.g., analyte) may be label free. For example, the sample may not comprise a label (e.g., a fluorophore) on an analyte. For example, the sample may not be coupled to a fluorophore (e.g., not covalently coupled to the label). In this example, the hyperspectral map can provide information related to the analyte without the use of a label. In some embodiments the sample without a label-may undergo de novo sequencing using the systems and methods disclosed herein.

[0198] In certain aspects, described herein is a method of determining a cellular response of a cell to a stimulus. Tire method may comprise on or more (e.g. up to all) of the steps of (a) providing a cell and a photonic chip, wherein the chip comprises a resonator; (b) attaching the cell to the photonic chip at or adjacent to the resonator; (c) exposing the cell to a first light; (d) measuring a second light emanating from tire cell; and (e) using at least the second light to identify the cellular response of the cell. The stimulus is applied to the cell before attachment to the photonic chip, or after attachment to the photonic chip. The second light may be a vibrational spectral signature, such as a Raman spectral signature or an infrared (IR) spectral signature. The method may be non -destructive to the cell. The method may be label free. The cell may be a mammalian cell, optionally a human cell. In some embodiments, the cell is a human T cell. Hie method may further comprise machine learning models trained to identify said cellular response of a cell based on a spectrum. Hie machine learning models may be trained to identify the cellular response of a T cell or state of a T cell based on Raman fingerprint interpretability determined by machine learning methods. In some embodiments, T cell differentiation or T cell sub-type is determined. The T cell differentiation may be selected from from CD4+, CD8+, thymic Treg, mucosal associated invariant T cell (MAIT), invariant natural killer T cell (iNKT), or induced Treg (iTreg). In some embodiments, the T cell sub-type is selected from Thl, Th2. Tel, Tc2, or Tnn. In addition to T cells, the method may be used for other immune cells (e.g., B cell, natural killer (NK) cell, neutrophil, eosinophil, basophil, mast cell, monocyte, or dendritic cell).

[0199] In certain aspects, described herein is a Raman macromolecule, comprising (a) a Raman reporter molecule; and (b) a barcoded peptide-MHC protein complex, optionally wherein the barcode is a DNA barcode. The Raman reporter molecule may have a distinguishable or identifiable Raman spectra based on the disclosure provided herein. A Raman reporter molecule as described herein may be an aromatic molecule having a unique Raman fingerprint. A barcode (e.g., a DNA barcode may have a distinguishable or identifiable Raman spectra based on the disclosure provided herein. In some embodiments, a barcoded MHC protein complex is may be a pMHC-dextramer, optionally wherein the barcoded peptide-MHC protein complex comprises a fluorophore. The pMHC-dextramer may be commercially sourced. The macromolecule may be for identifying components of a sample (e.g., an antigen).

[0200] In certain aspects, described herein is a method of producing a Raman macromolecule, the method comprising (a) providing each of (i) a Raman reporter molecule; and (ii) a barcoded peptide- MHC protein complex, and (b) attaching the particle and Raman reporter molecule of step (b) to the barcoded peptide-MHC protein complex. The particle may be a microbead or nanoparticle. The particle may be porous.

[0201] In certain aspects, described herein is a method of producing a Raman macromolecule, the method comprising (a) providing each of (i) a particle; (ii) a Raman reporter molecule; and (iii) a barcoded pcptidc-MHC protein complex; (b) attaching the particle to the Raman reporter molecule; and (c) attaching the particle and Raman reporter molecule of step (b) to the barcoded peptide-MHC protein complex. The particle may be a microbead or nanoparticle, optionally wherein the particle is porous.

[0202] A Raman reporter molecule as described herein may be an aromatic molecule having a unique Raman fingerprint. A barcode (e.g., a DNA barcode may have a distinguishable or identifiable Raman spectra based on the disclosure provided herein. In some embodiments, a barcoded MHC protein complex is may be a pMHC-dextramer. optionally wherein the barcoded peptide-MHC protein complex comprises a fluorophore. The pMHC-dextramer may be commercially sourced. The macromolecule may be for identifying components of a sample (e.g., an antigen). The particle may be attached to the Raman reporter molecule noncovalently, optionally wherein the noncovalent attachment is an intercalation of the Raman reporter molecule into the particle. The particle may be atached to the barcoded peptide-MHC protein complex with a cleavable linkage, optionally wherein the cleavable linkage is a disulfide linkage. The barcoded peptide-MHC protein complex may comprise a fluorophore.

[0203] A Raman macromolecule as described herein may be for use in vibrational spectroscopic determination of (a) a cellular response of a cell to a stimulus; (b) a cellular state of a cell; (c) a cellular type of a cell; or (d) a cellular binding activity of a cell. The method may further comprise machine learning models trained to identify7(a) the cellular response of the cell to the stimulus; (b) the cellular state of the cell; (c) the cellular type of the cell; or (d) the cellular binding activity of the cell based on Raman fingerprint interpretability determined by machine learning methods. Said machine learning models may be trained to demix the barcode identities when multiple barcodes are simultaneously present.

[0204] In certain aspects, described herein is a peptide-pulsed antigen presenting cell (APC), comprising (a) an APC; (b) at least one antigen, wherein each antigen is an immunogenic peptide; and (c) a Raman reporter molecule, wherein the Raman reporter is attached to the APC or to the antigen, and wherein the APC is pulsed with the at least one antigen. The Raman reporter molecule may be an aromatic molecule having a unique Raman fingerprint. The Raman reporter may be atached to the APC using a protein, peptide, lipid, or glycan.

[0205] The peptide-pulsed APC may be for use in vibrational spectroscopic determination of (a) a cellular response of an APC after stimulation with an antigen; (b) a cellular state of an APC after stimulation with an antigen; (c) a cellular type of an APC after stimulation with an antigen; or (d) acellular binding activity of APC after stimulation with an antigen. The method may comprise machine learning models trained to identify (a) the cellular response of the APC after stimulation with the antigen; (b) the cellular state of the APC after stimulation with the antigen; (c) the cellular type of the APC after stimulation with the antigen; or (d) the cellular binding activity of the APC after stimulation with the, based on Raman fingerprint interpretability determined by machine learning methods. The machine learning models may be trained to demix the barcode identities when multiple barcodes are simultaneously present.

[0206] In certain aspects, described herein is a method for determining an activated or unactivated state of CD8+ T cells using vibrational spectroscopy. The method may comprise one or more (e.g., up to all) of the steps of (a) contacting CD8+ T cells with a Raman macromolecule comprising a particle, and a Raman reporter molecule or a barcoded peptide-MHC protein complex, thereby producing a Raman macromolecule / T cell complex; (b) loading the Raman macromolecule / T cell complex into an array for vibrational spectra determination; and (c) determining the vibrational spectra for the Raman macromolecule / T cell complex at one or more timepoints; (d) immunostaining the T cells of the Raman macromolecule / T cell complex with an anti-CD137 binding molecule, and optionally staining the T cells of the Raman macromolecule / T cell complex with propidium iodide, and further optionally imaging the T cells of the Raman macromolecule / T cell complex using fluorescent microscopy; (e)performing fluorescence-activated cell sorting (FACS) on tire T cells of the Raman macromolecule / T cell complex based on anti-CD137 binding to sort for activated and unactivated CD8+ T cells, and optionally confirming results from the fluorescent microscopy imaging; and / or (f) perfonning single-cell RNA sequencing (scRNA-seq) to determine an activated or unactivated state of the T cells. The method may comprise machine learning models trained to identify (a) an activated or unactivated state of the T cells; (b) a cellular response of the T cells; (c) a cellular state of the T cells; (d) a cellular type of the T cells; or (e) a cellular binding activity of the T cells; based on Raman fingerprint interpretability determined by machine learning methods. The machine learning models may be trained to demix the barcode identities when multiple barcodes are simultaneously present. In addition to T cells, the method may be used for other immune cells (e.g., B cell, natural killer (NK) cell, neutrophil, eosinophil, basophil, mast cell, monocyte, or dendritic cell).

[0207] In certain aspects, described herein is a method for sequencing activated and exhausted CD8+ T cells using vibrational spectroscopy. The method may comprise one or more (e.g., up to all) of the steps of (a) contacting CD8+ T cells with a Raman macromolecule comprising a particle, and a Raman reporter molecule or a barcoded peptide-MHC protein complex, thereby producing a Raman macromolecule / T cell complex, wherein the Raman reporter molecule or the barcoded peptide-MHC protein complex isattached to the particle using a cleavable linkage, optionally wherein the cleavable linkage is a disulfide linkage; (b) loading the Raman macromolecule / T cell complex into an array for vibrational spectra determination; (c) determining the vibrational spectra for the Raman macromolecule / T cell complex at one or more timepoints; (d) immunostaining the T cells of the Raman macromolecule / T cell complex with anti-CD137. anti-PD-1, anti-LAG3, and / or anti-CTLA4 binding molecules, and optionally staining the T cells of the Raman macromolecule / T cell complex with propidium iodide, and further optionally imaging the T cells of the Raman macromolecule / T cell complex using fluorescent microscopy; (e) performing fluorescence-activated cell sorting (FACS) on the T cells of the Raman macromolecule / T cell complex based on anti-CD137, anti-PD-1, anti-LAG3, and / or anti-CTLA4 binding to sort for antigen-bound active and exhausted CD8+ T cells, and optionally confirming results from the fluorescent microscopy imaging; (f) cleaving the linkage between the particle and the Raman reporter molecule or barcoded peptide-MHC protein complex of the Raman macromolecule / T cell complex and isolating activated and exhausted antigen-specific T cells; (g) amplifying the TCR sequences from the activated and exhausted antigenspecific T cells; and / or (h) sequencing the TCR sequences from the activated and exhausted antigenspecific T cells. The method may comprise machine learning models trained to identify (a) an activated and / or exhausted state of the T cells; (b) a cellular response of the T cells; (c) a cellular state of the T cells; (d) a cellular type of the T cells; or (e) a cellular binding activity of the T cells, based on Raman fingerprint interpretability determined by machine learning methods. Timepoints for determining vibrational spectra of a sample may be selected from 10 minutes, 15 minutes, 30 minutes, 45 minutes, 1 hour, 2 hours, 3 hours, 4 hours, 5 hours, 6 hours, 10 hours, 12 hours, 18 hours, 24 hours, 2 days, or 3 days. Tire method may comprise machine learning models trained to demix tire barcode identities when multiple barcodes are simultaneously present. Tire method may comprise prior to step (a) the steps of (i) activating peripheral blood mononuclear cells with T cell activating peptides; and (ii) negatively selecting for CD8+ T cells. In addition to T cells, the method may be used for other immune cells (e.g., B cell, natural killer (NK) cell, neutrophil, eosinophil, basophil, mast cell, monocyte, or dendritic cell).

[0208] Any of the methods as described herein may be used in a high throughput and multiplexed screening of T cells and antigens. Any of the methods as described herein may be used in a high throughput and multiplexed screening of primary T cell activation from T cells, or peptide-pulsed antigen presenting cells. In addition to T cells, the methods may be used for other immune cells (e.g., B cell, natural killer (NK) cell, neutrophil, eosinophil, basophil, mast cell, monocyte, or dendritic cell).

[0209] Also provided are flow cells and / or well plates (e.g., nanowell flow cell) comprising a means for continuous tracking of individual T cells for at time points before, during, and after the introduction ofstimulants (e.g., pMHC multimers). Also provided are flow cells and / or well plates (e.g., nanowell flow cell) comprising a means for recovering T cells for sequencing after vibrational spectra determination.

[0210] Any of the methods described herein may have preceding steps comprising pre-processing the sample for vibration spectroscopy. Pre-processing may comprise any one or a combination of (a) chemical modification of the analyte; (b) fractioning the sample (e.g., isolating the analyte) for vibrational spectroscopy; (c) spotting or printing the sample for vibrational spectroscopy; (d) attaching the sample for vibrational spectroscopy to a photonic chip; (e) obtaining a vibrational spectra for the photonic chip (e.g., Raman spectroscopy); or (f) analyzing the vibrational spectra Data, optionally wherein the molecular species and of the sample and quantity of the sample is determined.

[0211] Sample pre-processing may comprise separating the sample into different fractions based on (a) solubility; (b) affinity pulldown; or (c) a resin based approach. Sample fractioning may comprise separating the sample into one or more fractions based on physical or chemical properties of the analyte, fractionization is performed using (a) liquid chromatography; (b) electrophoretic separation; (c) surface coated microparticles; (d) or a combination thereof. The physical or chemical properties may be selected from solubility, charge, amino acid composition, side chain group chemistry, post-translational modifications, size, or a combination thereof.

[0212] Chemical modification of the analyte may comprise modifying the analyte (e.g., peptide) at the N-terminus using benzaldehyde, 2-pyridinecarboxaldehyde, or a compound or resin comprising benzaldehyde or 2-pyridinecarboxaldehyde. Chemical modification of the analyte may comprise modifying the analyte at the N-tenninus using maleic anhydride, or a compound or resin comprising maleic anhydride. Chemical modification of the analyte may comprise modifying the analyte at the C- terminus using oxazolone formation at C-terminus, optionally by treatment with acetic anhydride and formic acid. Chemical modification may further comprise contact with an activator. The activator may be selected from pentafluorophenol, hydroxybenzotriazole (HOBt), hexafluorophosphate benzotriazole tetramethyl uronium (HBTU), O-(lH-6-chlorobenzotriazole-l-yl)-l,l,3,3-tetramethyluronium hexafluorophosphate (HCTU), or hexafluorophosphate azabenzotriazole tetramethyl uronium (HATU). Chemical modification of the analyte may comprise derivatization with an amine -containing compound. The amine— containing compound may be an alkyne-amine.

[0213] Chemical modification may comprise modifying a solid surface or substrate surface for covalent anchoring of an analyte with functional monolayers. The functional monolayers may comprise a surface reactive group linked to a peptide reactive group. Said surface reactive group may be selected from (a) asilane or silane group, optionally selected from triethoxy silane, tri methoxy silane, or chloro-silane; (b) a thiol based group; (c) a selenol-based group, and the peptide reactive group is selected from an NHS- ester, epoxide, carbodiimide, maleimide, or maleic anhydride. Chemical modification may comprise modifying a solid surface or substrate surface for covalent anchoring of an alkyne modified peptide molecule. Hie chemical modifications may comprise modifying the solid surface or substrate surface with one or more functional monolayers comprising a surface reactive group linked to an azide, optionally wherein the surface reactive group is a silane or silane group, thiol, or selenol. Chemical modification may comprise attaching an alkyne modified peptide to the azide modified surface, optionally via treatment with copper sulfate and sodium ascorbate.

[0214] Any of the methods described herein may have preceding steps comprising treating the solid surface or substrate surface to improve sample surface concentration. The surface treatment may comprise (a) oxygen plasma treatment; (b) alkyl-thiol monolayer deposition; (c) fluoro-alkyl -thiol monolayer deposition; (d) alkyl-silane monolayer deposition; (e) aluoro-alkyl-silane monolayer deposition; or (f) zwitterion-functional monolayer deposition, optionally selected from (sh-cl l-eg4- carboxybetaine or 3-{ [dimethyl(3-trimethoxysilyl)propyl]ammonio}propane-l-sulfonate).

[0215] Other linkage chemistries common for biomolecules (e.g., click chemistry) known to skilled artisan may be amendable as attachment means for analytes to chip surfaces as described herein.III. Computer systems

[0216] The present disclosure provides computer systems that are programmed to use any of the methods or systems as disclosed herein. FIG. 17 shows a computer system 1701 that is programmed or otherwise configured to implement the systems or methods of the present disclosure. The computer system 1701 can regulate various aspects of the present disclosure, such as. for example, detection and / or analysis of Raman spectra. The computer system 1701 can be an electronic device of a user or a computer system that is remotely located with respect to the electronic device. The electronic device can be a mobile electronic device.

[0217] The computer system 1701 includes a central processing unit (CPU. also “processor” and “computer processor” herein) 1705, which can be a single core or multi core processor, or a plurality of processors for parallel processing. The computer system 1701 also includes memory or memory location 1710 (e.g., random-access memory’, read-only memory’, flash memory), electronic storage unit 1715 (e.g., hard disk), communication interface 1720 (e.g., network adapter) for communicating with one or more other systems, and peripheral devices 1725, such as cache, other memory, data storage and / or electronicdisplay adapters. The memory 1710, storage unit 1715, interface 1720 and peripheral devices 1725 are in communication with the CPU 1705 through a communication bus (solid lines), such as a motherboard. The storage unit 1715 can be a data storage unit (or data repository) for storing data. The computer system 1701 can be operatively coupled to a computer network (“network”) 1730 with the aid of the communication interface 1720. Hie network 1730 can be the Internet, an internet and / or extranet, or an intranet and / or extranet that is in communication with the Internet. The network 1730 in some cases is a telecommunication and / or data network. The network 1730 can include one or more computer servers, which can enable distributed computing, such as cloud computing. The network 1730, in some cases with the aid of the computer system 1701, can implement a peer-to-peer network, which may enable devices coupled to the computer system 1701 to behave as a client or a server.

[0218] The CPU 1705 can execute a sequence of machine-readable instructions, which can be embodied in a program or software. The instructions may be stored in a memory location, such as the memory 1710. The instructions can be directed to the CPU 1705, which can subsequently program or otherwise configure the CPU 1705 to implement methods of the present disclosure. Examples of operations perfonned by the CPU 1705 can include fetch, decode, execute, and writeback.

[0219] Tire CPU 1705 can be part of a circuit, such as an integrated circuit. One or more other components of the system 1701 can be included in the circuit. In some cases, the circuit is an application specific integrated circuit (ASIC).

[0220] Tire storage unit 1715 can store files, such as drivers, libraries and saved programs. The storage unit 1715 can store user data, e.g., user preferences and user programs. The computer system 1701 in some cases can include one or more additional data storage units that are external to the computer system 1701, such as located on a remote server that is in communication with the computer system 1701 through an intranet or the Internet.

[0221] The computer system 1701 can communicate with one or more remote computer systems through the network 1730. For instance, the computer system 1701 can communicate with a remote computer system of a user. Examples of remote computer systems include personal computers (e.g., portable PC), slate or tablet PC’s (e.g., Apple® iPad, Samsung® Galaxy Tab), telephones, Smart phones (e.g., Apple® iPhone, Android-enabled device, Blackberry ®), or personal digital assistants. The user can access the computer system 1701 via the network 1730.

[0222] Methods as described herein can be implemented by way of machine (e.g., computer processor) executable code stored on an electronic storage location of the computer system 1701, such as, forexample, on the memory 1710 or electronic storage unit 1715. The machine executable or machine readable code can be provided in the form of software. During use, the code can be executed by the processor 1705. In some cases, the code can be retrieved from the storage unit 1715 and stored on the memory 1710 for ready access by the processor 1705. In some situations, tire electronic storage unit 1715 can be precluded, and machine-executable instructions are stored on memory 1710.

[0223] Tire code can be pre-compiled and configured for use with a machine having a processor adapted to execute the code, or can be compiled during runtime. The code can be supplied in a programming language that can be selected to enable tire code to execute in a pre-compiled or as-compiled fashion.

[0224] Aspects of the systems and methods provided herein, such as the computer system 1701, can be embodied in programming. Various aspects of the technology may be thought of as “products” or “articles of manufacture” typically in the fonn of machine (or processor) executable code and / or associated data that is carried on or embodied in a type of machine readable medium. Machineexecutable code can be stored on an electronic storage unit, such as memory (e.g., read-only memory, random-access memory, flash memory) or a hard disk. “Storage” type media can include any or all of the tangible memory of the computers, processors or the like, or associated modules thereof, such as various semiconductor memories, tape drives, disk drives and the like, which may provide non-transitory storage at any time for the software programming. All or portions of the software may at times be communicated through the Internet or various other telecommunication networks. Such communications, for example, may enable loading of the software from one computer or processor into another, for example, from a management server or host computer into the computer platform of an application server. Thus, another type of media that may bear the software elements includes optical, electrical, and electromagnetic waves, such as used across physical interfaces between local devices, through wired and optical landline networks and over various air-links. The physical elements that carry such waves, such as wired or wireless links, optical links, or the like, also may be considered as media bearing the software. As used herein, unless restricted to non-transitory, tangible “storage” media, terms such as computer or machine “readable medium” refer to any medium that participates in providing instructions to a processor for execution.

[0225] Hence, a machine-readable medium, such as computer-executable code, may take many forms, including but not limited to, a tangible storage medium, a carrier wave medium or physical transmission medium. Non-volatile storage media include, for example, optical or magnetic disks, such as any of the storage devices in any computers) or the like, such as may be used to implement the databases, etc. shown in the drawings. Volatile storage media include dynamic memory, such as main memory of such acomputer platform. Tangible transmission media include coaxial cables; copper wire and fiber optics, including the wires that comprise a bus within a computer system. Carrier-wave transmission media may take the form of electric or electromagnetic signals, or acoustic or light waves such as those generated during radio frequency (RF) and infrared (IR) data communications. Common forms of computer- readable media therefore include for example: a floppy disk, a flexible disk, hard disk, magnetic tape, any other magnetic medium, a CD-ROM, DVD or DVD-ROM, any other optical medium, punch cards paper tape, any other physical storage medium with patterns of holes, a RAM, a ROM, a PROM and EPROM, a FLASH-EPROM, any other memory chip or cartridge, a carrier wave transporting data or instructions, cables or links transporting such a carrier wave, or any other medium from which a computer may read programming code and / or data. Many of these forms of computer readable media may be involved in carrying one or more sequences of one or more instructions to a processor for execution.

[0226] Tire computer system 1701 can include or be in communication with an electronic display 1735 that comprises a user interface (UI) 1740 for providing, for example, control of Raman spectroscopy. Examples of UFs include, without limitation, a graphical user interface (GUI) and web-based user interface.

[0227] Methods and systems of the present disclosure can be implemented by way of one or more algorithms. An algorithm can be implemented by way of software upon execution by the central processing unit 1705. The algorithm can, for example, analyze the Raman spectra described elsewhere herein.

[0228] Exemplary peptide sequences for a method as described herein may be selected the sequence disclosed in any one of SEQ ID NOS: 1-19.EXAMPLES

[0229] The following examples are illustrative of certain systems and methods described herein and are not intended to be limiting.Example 1: Detection arrays

[0230] FIG. 12 is an example micrograph of a plurality of arrays, according to some embodiments. In this example, the scale bar is 200 micrometers. The plurality of arrays can be produced by, for example, lithography. In this example, the plurality of arrays has a density of three million arrays per square centimeter. Each array can be configured to detect an analyte. For example, all of the arrays can be configured to detect the same analyte. In another example, different arrays can be configured to detectdifferent analytes. FIG. 13 is a micrograph of an example array, according to some embodiments. The array can be configured as a guided mode resonance structure, which can be configured to concentrate light and / or control far field scattering. Tire array may comprise one or more photonic cry stal mirrors 1301 (the larger regions at the ends of the arrays). The photonic crystal mirrors can be configured to laterally confine a field (e.g., a field generated by incident light) into the guided mode resonant mode. In this way, far field incident light can be coupled into the modes of the array, which can then be concentrated as described elsewhere herein (e.g., via a gap present in a member of the array). FIGS. 14A - 14B are micrographs of example pluralities of arrays comprising a plurality of gaps, according to some embodiments. The arrays can be configured to concentrate the field of incident light within the gaps, thereby providing enhanced field strength as described elsewhere herein. FIGS. 18A-18D show additional examples of fabricated arrays, according to some embodiments. The arrays can comprise a pointed portion as in FIGS. 18C and 18D. The pointed portion can be configured to further increase the field in the nanogap. Similarly, FIGS. 22A - 22C show examples of fabricated arrays at different magnification levels, according to some embodiments. These arrays were fabricated with slot down the central axis of the array and with pointed portions to further enhance the field strength within the slot.

[0231] FIG. 1 shows an example of a processing workflow, according to some embodiments. A sample 101 can be introduced to a chip 100. The sample may comprise one or more molecules or components to be identified or detected. For example, a plurality of proteins can be suspended in a liquid and flowed into an inlet port 102 of the chip 100. The sample can be flowed through one or more features. The features can be as described elsewhere herein. For example, the features can be configured to filter the sample (e.g., by size, charge, etc.) to isolate analytes from the sample. The analytes can flow into the detection region 104, where the analytes can bind to capture probes 105, thereby placing the analytes into the near field of an array. The analytes may bind to the nanostructure and / or nanogaps as disclosed elsewhere herein. The array can then be configured to concentrate an incident light field as described elsewhere herein. Hie interactions of the analytes with the light field can produce a signal map 106. The signal map may be produced for each analyte in parallel. For example, the same incident light beam can interact with a plurality of arrays and a plurality of analytes to generate a plurality of spectra 107 at a same time. The spectra can then be analyzed to determine the identities of the analytes as described elsewhere herein.

[0232] FIG. 31 depicts an example of de novo peptide sequencing. In this depicted, an unknown peptide is coupled to a surface. The peptide undergoes amino acid cleavage and has each amino acid released individually and analyzed. The amino acid may be released from the protein using chemical methods such as for example Edman degradation or proteolytic cleavage. The released amino acid may be coupledto a nanogap. In some embodiments, the amino acid may be attached to the metal in the nanogap. In some embodiments, the peptide the amino acid is released from is attached to the metal in the nanogap. In some embodiments, the amino acid couples or attaches to the nanogap based on the chemistry of the nanogap as described elsewhere herein. The nanogap may comprise enough space for a single peptide or protein. A spectrum can then be gathered for the released amino acid as described elsewhere herein. The release and gathering of a spectrum for an amino acid can be repeated for the entirety of the amino acid sequence of the protein. Ultimately, the sequence and structure of the peptide is accurately predicted, including with the presence of post translational modifications.Example 2: Spectral analysis of analytes

[0233] FIG. 15 shows an example of spectral measurement of an interaction, according to some embodiments. In a first operation, the identity of an analyte can be determined. In this example, protein analytes are determined to be EGFR, JAK2, and an EGFR mutant (e.g., a unique proteoform of EGFR with a prost-translational modification at one or more amino acid residues). The identity can be determined using a Raman spectrum generated by an interaction of light with the analytes within an array, as described elsewhere herein. For example, each of the analyte proteins can be bound within a gap of an array, and light can be shone onto the array to determine the Raman spectra of the analytes.

[0234] A ligand can then be introduced to the analytes. For example, a ligand can be bound to the proteins. Tire binding dynamics can be tracked in real time via the Raman spectra of the analytes. For example, a time series of the analytes after introduction of the ligand can be taken, and the changes in the conformation of the analytes can be tracked via the spectra.

[0235] FIG. 16 shows sample Raman spectra, according to some embodiments. Tire Raman spectra can be generated by the methods and systems of the present disclosure. For example, an array comprising a gap can provide the Raman spectra. The arrays of the present disclosure may provide different degrees of field concentration depending on the polarity of the incident light.

[0236] FIGS. 19A - 19B show examples of Raman spectra of proteins and protein fragments according to some embodiments. The spectra may be generated from very small amounts of material using the arrays of the present disclosure. For instance, spectra may be generated from a single molecule using the arrays of the present disclosure. For instance, the spectra may be generated from about 24 attograms material using the arrays of the present disclosure. For instance, the spectra may be generated from less than about 24 attograms materials using the arrays of the present disclosure. Tire spectra may be generated from about two milligrams of material. In some embodiments, the material comprises asample. The differences between wild type and post-translationally modified mucin protein fragments may be discernible from the Raman spectrum of FIG. 19B. FIGS. 20A - 20B show examples of post translational modification spectra, according to some embodiments. FIG. 20B shows a plurality of Raman spectra of Ovalbumin epitope SIINFEKL (SEQ ID NO: 1) with different post-translational modifications. The spectra can be analyzed (e.g., via machine learning) to generate the scatter plot of FIG. 20A. The scatter plot can then be used to identify the post translational modification. FIG. 20C shows a normalized confusion matrix of the data of FIG. 20A, showing the high correlation between the predicted label and the true label in the sample.

[0237] FIG. 21 shows an example of a Raman emission versus excitation wavelength plot, according to some embodiments. The plot can show that the arrays of the present disclosure may be tuned to a predetermined resonance wavelength, which can provide a high intensity at the predetermined resonance wavelength. This can, in turn, enable multiplexing or multiple wavelengths to be directed to a chip comprising a plurality of arrays each configured to be resonant with a different illumination wavelength.

[0238] In some embodiments, the spectrum is compared to previously generated spectra. As a nonlimiting example, a spectrum is generated for each amino acid including with various PTMs. In the determination of an unknown protein, amino acids may be analyzed one at a time with a spectrum generated for each amino acid, such as indicated in Example 1. Tire spectrum generated for each amino acid may be compared against the spectra previously generated to determine the amino acid as well as any PTMs or other modifications associated with the amino acid. The combination of amino acids in sequence may lead to identification of the previously unknown protein.

[0239] The spectrum may also be developed for an entire peptide such as depicted in FIG. 30. In this depicted embodiment, a spectral catalog is built using experimental data and quantum calculations. Machine learning models are trained on the spectral catalog. When an unknown spectrum is generated for a peptide, the spectrum can be entered into the model trained on the spectral catalog to output the identity and structure of the peptide. In some embodiments, the process depicted in FIG. 30 and described herein may be used for proteins, amino acids, nucleotides, or any combination thereof

[0240] FIGS. 24A - 24B show examples of simulated clustered analysis of analytes and the associated simulated Raman spectra, according to some embodiments. FIG. 24 A shows an example of the strong correlation between the experimentally derived (e g., using the methods and systems of the present disclosure) spectrum and an ab-initio calculated spectrum. Using the same model, the Raman spectra of a plurality of different amino acids were simulated at varying degrees of noise and analyzed by a clusteringalgorithm as described elsewhere herein. As seen in FIG. 24B, the various amino acids show strong clustering when analyzed by tire algorithm, showing the ability of the methods and systems of the present disclosure to discern the different amino acids. Similarly, FIGS. 25A - 25B show7an example of a cluster analysis of a plurality of Raman spectra to identify analytes, according to some embodiments. Despite arginine-glycine-aspartic acid-serine (RGDS) (SEQ ID NO:2) and arginine-glycine-aspartic acid-cysteine (RGDC) (SEQ ID NO: 3) differing in composition by a single atom, the classifier was able to discern the difference with 100% accuracy. Similarly, both were identified as different from glycine-arginine- glycine-aspartic acid-serine (GRGDS) (SEQ ID NO:4).

[0241] FIGS. 26A - 26B show an example of Raman spectra and a difference spectrum associated with the introduction of a small molecule to a peptide, according to some embodiments. The topmost Raman spectrum of FIG. 26A shows a AYLGYLAML (SEQ ID NO:5) peptide alone, as noted in standard one letter code for each amino acid, while the bottom spectrum show7s the same peptide w ith a trifluoroacetic acid (TFA) and phenylisothiocyanate PITC additive. Hie changes in the molecular structure of the peptide result in the observed difference spectrum in FIG. 26B. This shows the ability of the methods and systems of the present disclosure to differentiate small changes in peptides or other analytes. A similar process can be used to discern, for example, protein-protein interactions, protein-ligand interactions, antibody-drug conjugate interactions, etc.

[0242] FIGS. 27A - 27D show an example of a confusion matrix and associated Raman spectra, according to some embodiments. FIGS. 27A - 27C show Raman spectra of three pairs of wild type versus single amino acid mutated major histocompatibility complex (MHC) peptides, as noted in standard one letter code for each amino acid, ATINFRRL (SEQ ID NO:6) vs ATINFRRR (SEQ ID NO:7), AYLGYLAML (SEQ ID NO:8) vs AYLRYLAML (SEQ ID NO:9), SCISKAML (SEQ ID NOTO) vs SSISKAML (SEQ ID NO: 11), respectively. The classification model was able to discern the various wild type and mutated with high accuracy, showing sensitivity to minor perturbations in the structure of the analytes.

[0243] FIGS. 28A - 28B show an example of a confusion matrix and associated Raman spectra, according to some embodiments. FIG. 28A shows a plurality of Raman spectra of various glycans as wild types, fucosylated, isomers, and with glycosidic linkages. The Raman spectra of FIG. 28A were processed using the clustering algorithms of the present disclosure, resulting in the confusion matrix of FIG. 28B. The confusion matrix shows highly accurate categorization of the glycans, showing the utility of the methods and systems of the present disclosure in glycan profiling.Example 3: Nanolithography process

[0244] The fabrication process for the nanolithography structures is developed on an area of about 15-40 mm2up 300 mm2called the write area. To start the fabrication process, alignment marks are deposited on the write area. As depicted in FIG. 32, the alignment marks comprise evenly spaced 25 micrometer (pm) gold or platinum squares along the perimeter of the write area and 5 pm gold or platinum squares along the interior. Tire base layer of the nanolithography structures comprises silicon deposited on top of another substrate such as, for example, silicon, glass, sapphire, another semiconductor, or another insulator. These base layers may be denoted by the acronyms SOS, SOG, SOS, or SOI, for example. The base layers can be deposited on the write area using an electron beam. For alignment marks, once the base layer is deposited on the write area with the electron beam, a layer of polymethyl methacrylate (PMMA) is added to the base layer using vapor deposition polymerization. PMMA used in this example comprises 950,000 MW resin in chlorobenzene with 4% solids. Hie PMMA and copolymer developer was used in a 1:3 ratio of methyl isobutyl ketone (MIBK) to isopropyl alcohol (IP A). This developer composition lends to very high resolution. The center of the nanolithography structure is targeted with dichloromethane (DCM) to remove some of the PMMA and create a channel. A four nanometer (nm) layer of chromium on which a 70 nm layer of gold or platinum may be deposited. Finally, DCM is used to remove the rest of the PMMA across the entire surface of the nanolithography structure removing the layer of chromium and gold or chromium and platinum that was deposited on top of the PMMA. This final structure for the alignment mark is the base layer with deposition of chromium / gold or chromium / platinum layer in the center when viewed aerially or when viewed from the side as the top layer covering only the center. These alignment marks are used for guidance during scanning electron microscopy.

[0245] Following the deposition of the alignment marks, hydrogen silsesquioxane (HSQ) is deposited on the base layer approximately 1 0 nm thick with the use of a developer comprising 1% sodium hydroxide and 4% weight sodium chloride. These are deposited on the write area generating a series of 30 pm pads. A portion of the HSQ and a portion of the upper silicon layer of tire base layer are etched using reactive ion etching. Hie reactive ion etching can be performed by hydrogen bromide-02 plasma and / or C 12-02 plasma. The etching of the top silicon layer with HSQ deposited on top produces a series of resonators within the 30 pm pads and result in a 50 nm gap between resonators (FIG. 33).

[0246] PMMA is deposited atop the HSQ and in the 50 nm gap (FIG. 34). A 15 nm section of the PMMA is removed from the center of the 50 nm gap. This section of PMMA can be removed using DCM as done previously for the alignment marks. A 4 nm titanium layer is deposited along with 10-20nm of gold across the PMMA surface as well as within the center of the nanogap. The remaining PMMA is removed from the surface leaving behind the resonators comprised of HSQ, silicon, and the bottom base layer along with a 15 nm disk comprising gold in the 50 nm gap between tire resonators wherein the 15 nm disk comprising gold is deposited directly on the bottom base layer substrate.Example 4: Use of resonators and nanostructures

[0247] Using the nanolithography process described in Example 3, a series of resonators comprising a nanostructure was generated. Analysis of the E field around the nanostructure with the Si / gold resonators resulted in enhancement of the E field (FIG. 35). Further, the location and number of the nanostructures deposited can be varied. As depicted in FIG. 36, a nanostructure within each gap surrounded by resonators. As depicted in FIG. 37, a nanostructure in the gap of only one of the six gaps shown. By varying the number of nanostructures deposited, the field enhancement location can be controlled. Additionally, the number of molecules in each cavity can be controlled by variation in the number and location of nanostructures.

[0248] FIG. 41A shows an example of a quality factor Q for a hybrid silicon / metal resonator, according to some embodiments. The high quality factor can provide improved signal gathered from the resonator, which can reduce the amount of analyte used to generate a given amount of signal. FIGS. 4 IB and 41C show properties of a plurality of resonators comprising various amounts of nanostructures, according to some embodiments . FIG. 41 B shows the peak values for the spectra of FIG. 41 C . Tire number of gold dot nanostructures present in the resonators can provide different amounts of signal enhancement to the resonator, as seen by tire dependence of the number of signal counts on the number of nanostructures. Even with a single nanostructure to provide field enhancement, detectable signals from a small number of analyte molecules can be detected. This demonstrates the ability of the resonator systems to receive signal from a small number of molecules using the combined resonators and nanostructures of the present disclosure.Example 5: Heterogeneous protein sequencing

[0249] FIG. 45 depicts sample processing and surface attachment of a peptide as described elsewhere herein. In this depicted embodiment, a peptide is reacted with N-(9- fluorenylmethylcarbonyloxy)succinimide and propargylamine. Tire reaction of the N- and C-termini of the peptide allows the peptide to attach in an oriented manner to tire surface of a nanogap. Additional examples of surface chemistries for attaching the peptide to the surface are depicted in FIG. 47. In some embodiments, the surface may be a nanostructure. In some embodiments, the peptide is part of aheterogeneous sample. The reactions and attachment as described for the one peptide above may be performed on all peptides in the heterogenous sample. The peptides may be attached to separate nanogaps. The peptides may be attached to separate nanostructures. The peptides may be attached to separate resonators. As depicted in FIG. 34, the peptides from a biological sample are individually attached to a metasurface. In this depicted embodiment, each metasurface is an individual resonator comprising nanostructures and one or more nanogaps. A laser is directed to each individual resonator allowing a Raman spectrum to be generated for each individual peptide. Processing of the peptide may be done using a machine learning model as depicted elsewhere herein.Example 6: Designing improved photonic chipsDual resonance features

[0250] Optimization of photonic chip design was pursued, with a goal of improving Raman spectroscopy sensitivity and sample processing throughput. In previous high quality factor (high-Q) chip designs, Raman enhancement comes from single high-Q resonance at the laser pump wavelength. However, single high-Q chips often require designs with larger length scales of resonant features to maintain large Raman enhancements. This limits tire ability to reduce device footprint for process scaling without significantly degrading performance.

[0021] A hybrid high-Q chip using both dielectric and antenna features was used to boost sensitivity and reduce device footprint. Silicon and metal were selected for the dielectric and antenna materials, respectively (exemplary resonances and surface electric field profile of a cichlid chip is shown in FIG. 48B (top) and the emission wavelength shown in FIG. 48B (bottom)). As a proof of principle, a hybrid high-Q resonator stack ~8 uM in length (FIG. 49A) with a metal dot at the center was designed (preliminary discus step) and emission (|E|2) of a single hybrid high-Q resonator was simulated. A graph of the | E|2electric field intensity line cut through the center (at peak resonance) is shown in FIG. 49B, showing that the enhancement of emission validated the approach of pursuing resonance synergy of chip design features. The total Raman enhancement effect is tire multiplied value of the electric field intensity at the excitation wavelength and the emission wavelength. For the discus designs the sensor only provides strong enhancement at the excitation wavelength while the newer cichlid design provides enhancement at both excitation and emission wavelengths leading to an overall stronger Raman signal.

[0252] A high-Q hybrid design having a non-resonant metal dot nanostructure produced strong simulated enhancement at the pump wavelength from the dielectric metasurface layer, but almost no enhancement from the metal only. FIGS. 50B— 50C overview the principle of resonance stacking, with dielectric, metalantennae, and hybrid overlays.. In this test, the overall design of the unit array was not perturbed, but high-Q resonance was enhanced and had better localization without degrading tire dielectric resonance.Dual-resonant designs engineer an antennae to provide a second dual resonance for further synergistic resonance effects (FIG. 50C). This approach differs from previous high-quality factor chip designs having metal features that fail to leverage the metal resonance for stronger functional purpose. Previous limitations were partially due to size constraints of metal nanostructures that can be accommodated onto chips, as well as strong metal quenching of the dielectric resonance that must be overcome.

[0253] A comparison of single-resonant and dual-resonant antennae designs is shown in FIGS. 51A. A dual-resonant hybrid design enables higher Raman enhancement than a singly-resonant hybrid design, and larger critical feature sizes. For example, if the minimum gap distance that can be reliably manufactured is about 20 nm (e.g., between two metal objects, or between a metal and a dielectric object). However, because of the synergy a dual-resonant design provides much larger Raman enhancementeven when using moderate quality factor resonances, compared to a singly-resonant design using high quality factor (FIG. 5 IB). The double resonance had improved Raman enhancement over varying minimum feature gap distances

[0254] The less strict size requirement allows for nanomanufacturing approaches not based on raster scanning (e.g.. electron beam lithography), which greatly improves the possible patterning efficiency and throughput. Dimensional scaling and patterning of unit cells on the chip can be optimized to improve performance over a broad range of wavelengths. Scaling dimensions are optimized for the wavelength bands of the photonic excitation source and the emission band. For example, an excitation laser band of -1064 nm and an emission band of -1120-1350 nm. or an excitation laser band of -785 nm and an emission band of -820-930 nm.Cichlid unit cell design

[0255] An improved unit cell pattern was designed containing a primary photonic pillar having one or more antennae composed of metal, which is surrounded by secondary photonic pillars lacking antennae. The unit cell is shown from the top view in FIG. 52A, which depicts a cichlid design having two antennae on a primary photonic pillar (at the center of the unit cell) and surrounding secondary photonic pillars. Although the cichlid design of FIG. 52A shows two metal antennae, different numbers of metal antennae are possible (e.g., 1, 2, 3, or more). Hie side view of the cichlid design is depicted in FIG. 52B and shows the layering of the substrate, dielectric material, dielectric spacer, and metal antennae. Common materials for the antennae include metals such as gold, silver, platinum, copper, aluminum, or titaniumnitride. Common materials for the dielectric include silicon, amorphous silicon, silicon carbide, silicon nitride, titanium dioxide, hafnium oxide, aluminum nitride, or gallium nitride. Common materials for the dielectric spacer include silicon dioxide, silicon nitride, aluminum oxide, or titanium dioxide. This cichlid design of FIGS. 52A-52B was manufactured by repeating unit cells across the substrate, as shown in the top views of chips in FIGS. 52C-52D. A dual antennae cichlid design boosted sensitivity (e.g., Raman enhancement of 10sor greater) and improved sample throughput since packing reaches 100M sensors per cm2,Improved laser line tolerance

[0256] In previous single high-Q only devices, Raman is enhanced by a strong but spectrally sharp resonance aligned with the incident laser wavelength. However, if the laser wavelength is slightly different than the high-Q pump resonance wavelength the Raman enhancements drop off dramatically, even with just 0.5 nm wavelength shift (as shown in FIG. 53 A for a laser having a wavelength of 1060 nm). Therefore, incredibly accurate fabrication is required with single high-Q only designs to ensure that all devices have the same resonance wavelength within <1 nm, or alternatively, requires the optical measurement system requires a more complex and expensive tunable laser to accommodate manufacturing defects.

[0257] In the hybrid mctal-diclcctric dual resonance system, a much lowcr-Q metal antenna based resonance at the laser pumping wavelengths can be utilized (for example, 785 nm as shown in FIG. 53B). To test laser line tolerance, Raman was measured on a cichlid chip design using a monolayer of thiolated molecules bound to gold antennae regions on the photonic pillar of a cichlid device (FIG. 53C). This low-Q resonance has a much broader spectral response, such that if the fabricated antenna resonance is shifted from the laser wavelength by plus or minus 10 nm, a large Raman enhancement can still be maintained from the device (FIG. 53D). This enables greater manufacturing tolerances in the fabrication of the devices. This also enables the optical measurement system to utilize much more stable, cost- effective, single wavelength laser sources as well.Improved optical loss tolerance

[0258] A device based solely on a high quality factor resonator suffers from significant performance degradation at low values of absorption or scattering loss represented by the complex permittivity (k) (see FIG. 54A). The hybrid cichlid device design was more robust to optical losses in the materials making up the device (see FIG. 54B). This also enables a broader range of materials to be used for the dielectric array, which can confer material processing and fabrication advantages.Improved coupling tolerance

[0259] A single resonance high-Q device is efficiently excited with a laser that is incident at exactly normal incidence (0 deg) is provided in FIG. 55A. Compared to the single resonance high-Q design, the hybrid resonance cichlid device was capable of more efficiently coupling incident illuminating light (see FIG. 55B).Flexibility of cichlid antennae placement

[0260] A comparison of cichlid vs discus chip designs is shown in FIG. 56. The cichlid approach enables geometrically flexible designs because it utilizes a non-gapped dielectric layer beneath the antennae. This allows for tighter-gapped structures because it does not rely on the manufacturing precision of the underlying silicon structure below for enhancement.

[0261] Tire use of a large silicon base in cichlid designs for metal layer lithography increases the reproducibility of resonances and eases manufacturing because placement of the antennae is less critical than when using gapped silicon. Antennae placement during lithography can vary (see FIG. 57A). However, each of these configurations produce nominally the same enhancement. As shown in FIGS. 57B-57C, vertical misalignments of 0 nm, 20 nm, and 40 nm have similar emission profiles. Additionally, the cichlid design is more tolerate of rotational misalignment of the antennae. This ease in alignment allows for manufacturing using high-throughput UV lithography, because there is less device to device variation in enhancement, which results in more repeatable data production.

[0262] The cichlid device is engineered such that equivalent field enhancement performance is obtained even when the upper layer is misaligned with respect to the lower layer. This opens up a broader range of device manufacturing options not typically used for nanomanufacturing in this regime, such as photolithography-based approaches which are more scalable across large areas but cannot guarantee exact placement of a feature on a prior fabricated structure to single digit nanometer accuracy.

[0263] Enhancement is achieved with relatively large silicon structures even at relatively short wavelengths, meaning that large metal structures (e.g., gold) can be accommodated on the second layer and still benefit from the dielectric enhancement. Tire ability to integrate large gold nanostructures allows a broader spectral range of dual resonances to be achieved, and improves the practical manufacturability of the feature dimensions. Greater flexibility in the range of metal shapes is possible (e.g. a doubletriangle bowtie structure with desired gap distance, which may not otherwise fit nicely with sufficient tolerance in other types of resonant dielectric structures).

[0264] Similarly, other modifications are possible, such as fabricating multiple metal nanoantennae per uniT cell, which can be useful if a higher number of sample capture sites on the chip is desired. Because a lower quality factor device is used to obtain Raman enhancement, the addition of more metallic structures does not degrade the resonator performance significantly.Adjustment of cichlid unit cell geometries

[0265] The cichlid device features stacked material layers (e.g., a dielectric layer and antenna layer) in which different resonances are engineered in each layer. In order to simplify tuning of feature dimensions to match the laser line and Raman band of interest simultaneously, an insulating spacer film was used to separate the dielectric and metal resonances, so that each could be tuned separately.

[0266] Tire dimensions of features in the lower layer (e.g., dielectric) of the cichlid design are tuned for the size of the photonic pillar, as well as the width of the unit cells. Tire field enhancement effects (|E| / |E0|) of primary photonic pillars having disk radii of 120 nm and 135 nm are shown in FIG. 58A, where a silicon dielectric was tested with gold antennae. The field enhancement effects (|E| / |E0|) of different sized unit cells for this design with total widths of 500 nm and 575 nm are shown in FIG. 58B.

[0267] The upper layer (e.g., antennae) of the cichlid design was tuned separately by adjusting dimensional geometries. The overall size of antennae is limited by the diameter of the silicon beneath it. A maximum diameter of about 300 nm is typical for metal antennae. However, the antennae dimensions can vary widely depending on the geometric shape, material, thickness, and the number of antennae.

[0268] The field enhancement effects (|E| / |E0|) of antennae bowtie lengths of 70 nm and 80 nm with a silicon dielectric and gold antennae is shown in FIG. 58C. Typical dimensional measurements for antennae shapes are shown in FIG. 59, and TABLE 1 below. Additional simulations and experiments for separate tuning of disc radius and unit cell height are shown in FIGS. 60A-60D.

[0269] TABLE 1: Antennae dimensions

[0270] Example validation data (FIGS. 61A-61F) showed optimization of excitation / emission wavelength for co-designing multi-band chip enhancement. A heatmap showing Raman intensity for various bowtie lengths and disc radii at a 785 nm pump wavelength is provided in FIG. 61 A. FIG. 6 IB shows reporter Raman for a fixed disk radius (variation in emission wavelength enhancement due to Si resonances), with varying pump wavelengths and bowtie lengths. Strong Raman enhancement can be obtained across a range of pump wavelengths on a single device, allowing flexibility in choice of laser and detector. Data from a 1060 nm pump range for varying bowtie lengths and disk radii is shown in FIGS. 61C-61F. In general, the largest Raman signal came from optimized spectral overlap of the two tuned modesImproved photonic pillar densities and reducing sensor crosstalk

[0271] Simulations were performed on a cichlid chip design to determine the level of crosstalk between photonic pillars when the sensor densities is >100 M / cmA2. Results are shown in FIGS. 62A-62D. Excitation crosstalk was shown to be < 1% for diagonal nearest neighbors, and <0. 1% for horizontal nearest neighbors.

[0272] Additional tests were performed for a photonic pillar density of ~400 M / cmA2. When the polarization of the sensors were aligned (and spaced 500 nm apart), crosstalk up to 15% was observed (see FIGS. 63A-63C). It was found that a reduction in crosstalk could be achieved by rotating the polarization of adjacent sensors. With the same 500 nm spacing, a 90 degree rotation of the sensors reduced crosstalk to 0.5% (see FIGS. 63D-63F). From these results, it was determined that antennae are polarization dependent, and that cross-polarization of antennae on neighboring photonic pillars enables denser packing while reducing cross-talk.Alternative unit cell designs

[0273] Using the design principles gleaned from results of the initial cichlid designs (e.g., bottom of FIG. 56), additional antenna geometries suitable for multi -resonance high-Q chips were developed. Designs having one or multiple antennae and / or antennae having different geometric shapes were designed (see FIG. 64).

[0274] Geometries of the antennae included triangular, rectangular, circular, elliptical, and other geometric shapes. Shapes that feature linear edges (e.g., polygons such as triangles, rectangles, pentagons, hexagons, heptagons, and octagons) are generally well suited for cichlid chip designs, due totheir ease of manufacture compared to elliptical or other shapes lacking linear edges. Polygon-based silicon resonator designs (such as those shown in FIG. 65) enable more uniform and conformal oxide filling of dielectric (e.g., silicon) plane by having more uniform gaps between silicon features. This aids chemical mechanical polishing (CMP) for multi-layer fabrication (e g. of the subsequent layers or antennae), and helps reduce non-specific or stray enhancement from the dielectric resonator outside the antennae sensing area.

[0275] Further cichlid designs having varying unit cell arrangements were also developed. These approaches adjusted the density of secondary photonic pillars (e.g., those lacking antennae), as well as the orientation (e.g.. polarity) of the antennae on the primary photonic pillar. Suitable antenna 2D spatial arrangements are shown in FIG. 66. Additional cichlid designs tailored for spatial multiplexing (e.g., pump wavelength control) were also designed (see FIG. 67).

[0276] The initial cichlid layering design (e.g., schematic of FIG. 52B) was also improved. FIG. 68A shows a schematic of a hybrid design that incorporates a dielectric layer (e.g., silicon), an oxide layer, a silicon layer, a dielectric spacer, and antennae. FIG. 68B shows a schematic of a hybrid design that incorporates a metal layer, an oxide layer, a silicon layer, a dielectric spacer, and antennae. FIG. 68C shows a schematic of a hybrid design that incorporates a substrate layer, a silicon layer, a dielectric fill layer, and antennae. FIG. 68D shows a schematic of a hybrid design that incorporates a substrate layer, a silicon layer, a dielectric spacer, a passivation layer (to isolate enhancement spots and reduce background enhancement from bulk Si surfaces) and antennae.Minor-containing cichlid designs

[0277] FIG. 69A shows a schematic of a mirror-enhanced hybrid chip design, which incorporates a metal or dielectric mirror layer, a dielectric spacer layer, a silicon layer, a dielectric fill layer, and antennae. Generally, the mirror structures may be composed of a metal layer, a metal coated silicon or glass wafer, or a dielectric mirror comprising alternating layers of silicon and oxide on a silicon wafer. Incorporation of a mirror into the design resulted in Fabry -Perot increase of the enhancement factor due to interference effects, due to increased excitation efficiency at pump wavelength and increased emission photon collection and detection (see FIGS. 69B-69C). Additional variations of mirror enhanced chips are shown in FIGS. 69D-69E. FIG. 69D shows a schematic of a mirror-enhanced hybrid chip design that incorporates a silicon or glass wafer layer, a metal layer, a dielectric spacer layer, a silicon layer, a dielectric fill layer, and antennae. FIG. 69E shows a schematic of a mirror-enhanced hybrid chip design that incorporates a silicon or glass wafer layer, alternating layers of silicon and oxide layers, a dielectricspacer, a silicon layer, a dielectric fill layer, and antennae. Further simulations varying the thickness of the supporting layers (e.g., oxide thickness) in mirror-containing designs showed improved performance for certain pump wavelengths (see FIGS. 69F-69H).Example 7: Fabrication of cichlid chips

[0278] The manufacture of photonic chips as described herein was performed using deposition and lithography processes to yield resonant enhanced devices for use in vibrational spectroscopy (e.g., Raman spectroscopy). Mirror enhanced cichlid chips were manufactured according to TABLE 2. A graphical overview of the manufacturing steps is provided in FIGS. 70A-70Q.

[0279] TABLE 2: Overview of chip fabrication

[0280] The above manufacturing approaches produce a dual-resonant cichlid chip design, but may be adapted to suit other non-cichlid designs disclosed herein.Example 8: Improving chip reads

[0281] Performing high throughput Raman spectral readout on an array of cichlid devices (or any patterned SERS substrate where the hot spot locations are predefined) has several challenges. One challenge is that there must be single-device spatial resolution. Another challenge is that adequate signal must be collected from a weak interaction (Raman scattering), without degrading the sample. Also, there must not be cross-talk between emissions of different devices on a chip.

[0282] Prior approaches to reading out arrays with microscope modalities have certain disadvantages. For instance, a widefield technique uses low irradiance on the device, which results in low signal. Most of the illumination power is wasted illuminating the area between devices. Furthermore, it is difficult to couple emission to spectrometer, and emission cross-talk is typically present between closely spaced devices. A confocal technique using a single spot has extremely high irradiance, which leads to photo and thennal degradation of the sample. Additionally, depending on confocal pinhole size, the emission region may not be fully captured, which results in a low signal.

[0283] Therefore, improved methods for reading chips were developed that combines widefield and confocal techniques. Briefly, an array of illumination spots efficiently concentrates the available power on devices but also spreads out the irradiance over multiple spots to avoid sample degradation and crosstalk. An array of detection regions concentric with the illumination spots are used to collect emission,where each detection region is large enough to collect emission in the presence of spreading. A graphical overview of the chip reading technique is shown in FIG. 71.Laser source

[0284] At least one laser source is utilized. The laser is suitable for Raman spectroscopy, having narrow linewidth (e.g., linewidth of < 10 cm1, such as 5 cm1or 2 cm1), high side mode suppression, low wavelength instability, high spatial coherence (M2 < 2), and high degree of polarization. Multiple lasers can be used.Excitation spot array generation

[0285] A one -dimensional or two-dimensional array of spots may used. For better performance, the arrays can tile the plane to make the scan simple. For example, a rectilinear array is used.

[0286] Methods for spot generation are diffractive optical element (DOE), microlens array, and fiber splitters / array. The DOE technique is preferred because of its low loss and low complexity optics.

[0287] Tire spot pitch should be large enough that the emission region and thermal influence region from nearest neighbor illuminated devices do not overlap. When the number of spots is increased, the irradiance at each spot decreases, which lowers the risk of sample degradation. However, the distance between spots decreases because the whole array must fit within the objective’s field of view. Eventually the pitch decreases so much that the emission regions may overlap. A large number of spots also makes the fiber bundle used for tire Relay very expensive, because a large number of fibers must be precision assembled into the bundle.Sample plane

[0288] Outer dimension are defined by objective image circle. For lOOx objective, about 125 x 125 um. For 50x objective, about 250x250 um. Other objectives should match these ratios as a starting point.Excitation spot (ij)

[0289] Spots can be diffraction limited in diameter, though this is not required. Diffraction limited is the best case because it gets the highest irradiance on the device for a given source power. Generally speaking, a high numerical aperture objective should be used for highest irradiance and highest light collection.Thermal influence region for spot (ij)

[0290] Some of the excitation is absorbed in the substrate causing a temperature increase. If an excitation spot is placed within the thermal influence region of a nearby spot, then the temperature will increase much more than if the excitation spots were more widely spaced. In general, excitation spots should be spaced as far apart as possible.Emission region for spot (i,j)

[0291] Due to the resonant structure, devices do not necessarily emit as point emitters. The emission region is tire patch of the surface from which emission from a given device occurs. In ideal situations, the emission regions from illuminated devices do not overlap, or else it would be very difficult to unmix the combined emission.Field of view corresponding to spectrometer track (k)

[0292] Light emitted within the field of view (z,j) is collected and relayed into track (£) of the spectral readout (see FIG. 71). In the best case, this field of view is at least as large as the emission region so that all the emission is collected.Spectrometer

[0293] The spectral readout is based on multitrack spectroscopy. Therefore, an imaging spectrometer is required, c.g. Schmidt-Czerny -Turner.Scan

[0294] The scan translates the spot array relative to the device array, moving all spots at once by the same amount. The scan can be accomplished using a mechanical motion of the sample stage or a galvanometers mirror scan of the spot array. The best solution is a mechanical motion of the sample stage because it's simpler and more robust.Relay

[0295] The relay is an optical system that maps emission from the array of spots (z,j) onto a linear array of spots (k) on the spectrometer entrance slit (see FIG. 71). Each spot is dispersed into a track (£) on the spectrometer detector, which are digitized into spectra. Ultimately, each spectrum (L) corresponds to the emission of one illuminated device (i,j)

[0296] Tire relay can be free space, if the spectrometer is placed with its entrance slit at the focal plane of a microscope. At least one lens could also be used to relay. The preferred relay solution is a fiber bundle,which can reshape a 2D array of spots to a ID array for coupling into the spectrometer. In this case, a microlens array may be coupled to the fiber bundle to increase the amount of light captured. The number of fibers and fiber numerical aperture should be chosen to capture all emission from the devices while still fitting within the spectrometer aperture, limited by conservation of etendue.Example 9: Dynamic monitoring of single cell responses

[0297] A label-free method for dynamic monitoring of single cell responses to stimuli was developed. Initial development was performed using human T cell, however other cell types may be used. T cells are a crucial component of the adaptive immune system, playing an important role in identifying and eliminating pathogens, recognizing and attacking cancerous cells, and discriminating self from non-self. Their ability to adapt and respond to a wide array of threats is largely attributed to their T cell receptor (TCR) diversity, which enables recognition of various antigens to mount tailored immune responses. T cells are induced into new states through binding of their TCRs to antigens presented on antigen- presenting cells (APCs), accompanied by the binding of other cell surface co-stimulatory factors such as the T cell’s CD28. This attachment between the T cell and APC produces a cascade of intracellular signaling, leading to changes in a T cell's functional state by modifying transcriptional programs and cytoskeleton rearrangement, among other changes. This adaptive immune response targets APCs for death. The complex transitions between states (e.g.. activated, memory, effector, exhausted, and many gradations in between) are nuanced and result in widely varied speed and efficacy against threats.Understanding the activity, and dysfunction, of T cells is crucial for engineering effective therapeutics that leverage T cells against a huge range of pathologic conditions, including infectious disease, cancer, and autoimmune disorders.

[0298] The currently available technologies for probing T cell states against antigens pose a challenging dichotomy for researchers; experimental options are generally either highly detailed but static, or dynamic but low resolution. Among the exquisitely detailed but static methods are flow cytometry, cytometry by time of flight (CyTOF), and single cell sequencing. Flow cytometry’ and CyTOF define T cell states by cell surface markers and intracellular cytokines with -8-12 (flow) to 25-50 (CyTOF) parameters per experiment. High frequency of sample collection is often precluded by cost (~$2-10k / sample), logistical challenges, or patients and blood volume for research animal models. Importantly, while floyv cytometry and CyTOF can analyze antigen-specific responses through the introduction of antigens (peptide-MHC tetramer constructs), often only a few (-1-25) antigens can be analyzed at once due to spectral overlap of fluorescent / metal markers required for detection. Single cell RNA sequencing (scRNAseq) complements these methods through the analysis of transcriptional programs. A technique called “pseudotime”analysis was developed in an effort to uncover more dynamic insights from single time point scRNAseq data. However, the ability’ to infer dynamics from a single time point, or even a handful of time points, leans heavily on significant assumptions, potentially leading to significant bias in results and limiting reliability.

[0299] At the other end of the spectrum are methods that deliver more dynamic information, but for a general T cell population, obscuring the function of individual cells / TCRs. These methods include experiments run on bulk cell populations such as T cell proliferation and cytotoxicity assays, and cytokine production monitoring (ELISpot). In these experiments, ~ 10k- IM T cells are stimulated by a single antigen (or a pool of related antigens) to observe a single functional output. In a proliferation assay, initial T cells are incubated with a marker such as CFSE, or BrdU that serves as an indicator of general cell division as is expected upon stimulation of a T cell by its TCR. ELISpot assays report stimulation through enzyme-linked immunosorbent (akin to ELISA assays) reporting of cytokine production (a single, specific cytokine, often IFN-gamma is detected). While each of these techniques offer important insights into T cell biology, there is a critical gap in our ability to understand dynamic T cell function on individual cells, and across many (e.g., hundreds to thousands of) antigens.

[0300] Human leukocyte antigens (HLAs) are proteins that bind to peptide fragments from antigens and present them to T cells. HLAs are the human version of major histocompatibility complex (MHC) molecules. T cell receptors (TCRs) are protein complexes on the surface of T cells that recognize antigens presented by HLAs. HLA-TCR interactions are a key step in the immune response, helping the body identify and eliminate unhealthy cells. There is enormous diversity in T Cell-Antigen interactions. Humans generally have 3 to 6 HLA Class I and 3 to 6 Class II HLA alleles. Overall, there are over 27,700 different Class I and 11,900 different Class II HLA alleles in the human population, but only a small handful are well studied (Hudson et al., Nature Reviews Immunology (2023)). There are an estimated ~209possible antigens (Lin et al., Nature Compututational Science (2021)). Possible antigens comprise infective agents, such as viruses and bacteria, but the relevant antigen set will differ by patient HLA type. Other types of antigens include cancer, in which neoantigens are generally unique to an individual tumor (or patient). Each person has ~5M unique TCRs, and about ~11% of each person’s TCRs are shared with others. At the population level, there are about 1015- IO20unique TCRs in the human population. Each TCR binds to -10.000-1M self peptides with varying affinity (Hudson et al., Nature Reviews Immunology (2023); Lin et al.. Nature Compututational Science (2021)).

[0301] Relative to the scope of the problem, very few TCR-antigen pairs are known. The number of TCRs per antigen species of origin has been previously investigated, and the majority of all antigensreported as binding a TCR are of viral origin. Moreover, a group of 100 antigens makes up 70% of known TCR-antigen pairs. These known antigens are reported in complex with only a few common HLA alleles. Most antigens have only one known cognate TCR in the combined data set (Hudson et al., Nature Reviews Immunology (2023)). There are likely more human / cancer pairs that are described in non-public repositories (e.g., proprietary to drug discoven’ companies).

[0302] Current data sources have a few problems: 1) pre-selection with pMHC models, and 2) low- sensitivity immunogenicity screens. Screening preselection with pMHC presentation models relies on accuracy of a decision boundary, and immunogenic antigens may be mistakenly filtered out of pMHC selection. Immunogenicity screening experiments may have false negatives due to low screening sensitivity.

[0303] The label-free methods described herein (e.g.. a Raman-based deSIPHR platform; see FIG. 72) provide means for fingerprinting and / or sequencing biological molecules such as protein or nucleic acids. This approach is now extended to whole cells, in a cellMAPP assay described herein. Using Raman- labeled microbeads (i.e., a “barcode”) and pMHC multimers (e.g., pMHC dextramers), a multiplexed antigen library are contacted with a repertoire of T cells and the complexes distributed in a nanowell array (e.g., enabling isolation of single cells). Vibrational spectra is then determined (e.g., Raman spectroscopy). Spectra for cell signatures and antigen barcodes are determined, and optionally applied to a machine learning model trained for identification of the spectra. This approach makes possible the monitoring of live cells (see FIG. 73).

[0304] A graphical explanation of the Raman barcoding is provided in FIG. 75. A unique Raman reporter, made of small aromatic molecules that present sharp but distinct Raman peaks, serves as the barcode that distinguishes between each antigen. The polystyrene core of microbeads are soaked in organic solvents solutions of reporter, causing the core to swell and reporter molecules to intercalate. After rinsing of the particles in aqueous solution, the cores to shrink and trap reporter molecules inside the core. Due to the sharp and non-overlapping spectral features in Raman spectra, Raman barcoding theoretically scales to thousands or even a million unique signatures for high throughput screening. Tire Raman barcoded cores are functionalized with pMHC -dextramers that are labeled with an antigen specific DNA barcode (e.g., off-the-shelf from Immudex designed for compatibility with 10X Genomics Chromium System; numerous HLA alleles available), such that antigen binding and cell state trajectories are linked to subsequent single cell sequencing.

[0305] The antigen presenting construct are modified such that the pMHC-dextramers also host a fluorescent label and are attached to the microbead core with a cleavable disulfide linker (FIG. 75). Following T cell monitoring, a reducing agent (e.g., tris(2-carboxycthyl)phosphinc (TCEP)) is used to reduce the disulfide linkage. The microbeads are then be removed, while the dye labeled pMHC- dextramer remains bound to the cell.

[0306] CellMAPP produces data including antigen-TCR sequence pairings, as well as cell state populations with time resolution. Various antigens can be contacted with the T cells to determine unactivated, activated, and exhausted states. The types of TCR-HLA (antigen) pairs are determined (see FIG. 74). cellMAPP provides the ability to multiplex large antigen libraries while maintaining TCR specificity. A comparison of cellMAPP with current T cell analysis technology is shown in FIG. 76, showing that cellMAPP has advantages of providing dynamic observation, with high T cell resolution, high antigen throughout, in a cost-effective TCR specific assay.

[0307] The cellMAPP workflow for capturing Raman fingerprints of unactivated and activated CD8+ T cells is shown in FIGS. 78. Briefly, starting with peripheral blood mononuclear cells (PBMCs), the PBMCs arc preactivated with peptides over a scries of days. Then a negative selection of CD8+ T cells is performed. Then the CD8+ T cells are mixed with pMHC dextramer beads for reactivation. The beads and T cells are loaded into sievewell nanowell array. Next, Raman fingerprints are collected for all T cells at different time points (e.g., 0, 1, 3, 6, and 12 hrs). The cells are then immunostained with an appropriate antibody (e.g., anti-CD137 for CD8+ T cells) and a cell stain (e.g., propidium iodide) in a nanowell array and imaged using fluorescent microscopy. The stained cells are then recovered and fluorescence-activated cell sorting (FACS) is used to confirm fluorescent microscopy data. Sequencing is perfonned (e.g., scRNAseq) to confirm activated and unactivated cell states.

[0308] Tire cellMAPP workflow for recovery of activated and exhausted CD8+ T cells for TCRSeq is provided in FIG. 79. Tire method deviates from the fingerprinting assay above at the staining step. Cells are instead immunostained with anti-CD137, anti-PDl, anti-LAG3, and anti-CTLA4 antibodies, and stained with propidium iodide in the nanowell array and imaged using fluorescent microscopy. Hie stained cells are then recovered and FACS is performed to confirm the fluorescent miscropy data, and to sort antigen-bound activated and exhausted CD8+ T cells. A reducing agent is then added to separate beads from cells, while retaining TCR / MHC interaction. The antigen-specific T cells are loaded into 10X Genomics Chromium instrument to amplify TCR sequences. Finally, the amplified TCRs are sequences using an Illumina NovaSeq system.

[0309] Strategies for coupling Raman reporters to antigen presenting cells is provided in FIG. 80. The APC may be coupled to the Raman reporter using a protein, lipid, or glycan moiety.

[0310] The cell-MAPP techniques provided herein enable an unprecedented, dynamic monitoring of live individual T cells throughout their respective trajectories upon stimulation. Antigen barcoding and post- cell-MAPP scTCRseq enable pairing the T cell responses to a specific antigen-TCR pair. Cell-MAPP offers 3 key capabilities for T-cell screening: 1) single-cell, dynamic monitoring of T-cell states upon antigen stimulation 2) high-resolution insight into cell responses and 3) high throughput analysis through highly multiplexed individually-identifiable antigen stimulation (>l,000s antigens on >1M single cells).

[0311] These advances allow researchers and therapeutic developers novel insights into key immune system detenninants of diseases and therapeutic response. Cell-MAPP evaluates stimulatory capacity among a set of antigens under evaluation for inclusion in infectious disease or cancer (personalized or off- the-shelf) vaccines. Cell-MAPP can also probe the T cell modulating capabilities of different drug strategies, including checkpoint inhibitors (which unleash T cell cytotoxicity through blunting T cell inhibitory pathways) or T cell engagers (which can physically link T cells to cancer cells via CD3 as well as stimulate them). Finally, cell-MAPP could monitor the therapeutic capacity of engineered CAR-T cell therapies to ensure persistent, cytotoxic behavior against target antigens.Example 10: Improved proteomics workflow

[0312] A comparison a traditional mass spectrometry' (MS) proteomics workflow and a vibrational spectroscopy-based proteomics workflow is depicted in FIG. 81. Both approaches involve extraction and testing of peptides (c.g., from a cell of interest). In the traditional MS workflow, the extracted peptides are fractionated by HPLC, then ionized and detected on the mass spectrometer. Due to sample loss, improper ionization, and weaker detection limit, the sensitivity the MS approach can be problematic.

[0313] However, vibrational spectroscopy as described herein may be utilized instead. In a proteomics workflow based on this approach, following HPLC fractionation of the peptides, the sample is loaded onto a chip (which has lower sample loss than ionization). The improvement in sample loss is partially due to the ability to utilize nanoliter volumes of liquid for the technique. A sensor readout of the vibrational spectra is performed, which has a lower limit of detection. Compared to a MS approach, the vibrational spectra approach (e.g., Raman spectroscopy) has an improved limit of detection by approximately 3-4 orders of magnitude.

[0314] General flow charts of key steps in the vibrational spectroscopy (e.g., Raman spectroscopy) proteomics approach are shown in FIGS. 82A-82C. Briefly, the approach involves sample preprocessing; optionally chemical modifications of peptides; peptide mix separation (e.g., HPLC); printing peptide fractions on an engineered substrate (e.g., a photonic chip as described herein); optionally improving surface chemical attachment; performing vibrational spectroscopy (e.g., Raman spectroscopy); and analyzing the data.Sample pre-processing

[0315] Tire sample pre-processing steps involve isolating the desired peptide sample, such as by purifying based on solubility (e.g., water vs water / ACN / TFA or other solvent mixtures), size and sequence vs. charge (e.g., isoelectric point); affinity purification; reverse phase separation; electrophoretic mobility separation, a resin-based separation (e.g., zip-tip); other chromatographic (e.g., HPLC) technique. The choice of solvent is selected based on the properties of the peptides at optimum concentrations and volumes. Examples of solvents include pure water, which may dissolve 50% or more of peptides in a mixture; or water / ACN / TFA, which may dissolve 90% peptides in a mixture.Peptide separation

[0316] After sample dissolution and pre-processing, the sample is further isolated or separated using chromatography. Hie sample peptide mixture is injected into an instrument for separation (e.g., using an HPLC, or capillary electrophoresis). Separation parameters are determined for optimal fractionation. Peptide separation is monitored by absorbance or fluorescence on the instrument giving a chromatogram trace, which can be used to deduce information about the peptide fraction. Peptide fractions are collected (e.g., into well plates). Peptide fractions are linked with an interface for printing onto the sensor (e.g., a photonic chip as described herein). Generally, multiple steps of isolation and fractionation are performed. For example, a first fractionation of a >10k peptide mix cuts into smaller fractions containing -100 peptides or less. Additional fractionation steps are performed (e.g., to cut down to <10 peptides).Multiple fractionations are performed to produce a fraction of peptides having desired physical properties.

[0317] Data for a multi-fractionation are shown in FIGS. 83A-83C. FIG. 83A shows sample preprocessing by solubility followed by peptide fractionation by HPLC. A chromatogram trace showing sequential dissolution and separation of 138-peptide pool into fractions with different distribution of peptides is provided. Next the sample was run through the HPLC under the same method (C 18 column, ACN / water / TFA gradient) over 2h. A different distribution of peaks was observed between the two approaches. In the first cut (top of the graph), the water soluble peptides elute first. The aqueous samplethus contains a majority of peaks since it was performed first, and most of the soluble peptides are eluted. The second spectrum in FIG. 83A (bottom of the graph) for water / ACN / TFA solvent shows peptides undissolved in tire first run.

[0318] FIG. 83B shows data for peptide fractionation and separation of a target spiked in peptide peak by HPLC. The bottom of FIG. 83B shows a mixture of 800+ peptides, which produced approximately 80 peaks over a 40 minute run using a C 18 column and ACN / water / TFA gradient. The top of FIG. 83B shows the same sample, which was spiked with a target protein. The asterisks indicates where the target protein appears in the chromatograph.

[0319] FIG. 83C shows data for the effect of solvent gradient on separation of mixture using a lesser number of peptides per fraction. The separation of peaks was improved upon adjustment of the gradient. The bottom portion of FIG. 83C shows the faster (1 hour) gradient, and the top portion of FIG. 83C shows the slower (2 hour) gradient. The sample contained 138 peptide pooled and was separated using a C18 column with an ACN / water / TFA gradient mobile phase.Improved attachment to photonic chips

[0320] Various attachment chemistries for attaching biological molecules (e.g.. peptides, proteins, nucleic acids, etc.) was developed. Generally, any attachment mechanism can be used as long as the spectral properties of the biological molecule remain identifiable. Attachment techniques utilized include DSP, NHS ester, GLYMO, EDC, carbodiimide, maleimide, and maleic anhydride based coupling (see FIG. 84. Alternatively, surface anchoring may be used (e.g., thiol, selenol, silane, and chlorosilane) attachment; see FIG. 85). The attachment is generally performed at the end of the molecule (e.g., N- tenninal modification; C-tenninal modification; 5 ’-modification; 3 ’-modification; etc ). Alternatively, a bead based reversible amine attachment may be used (e.g., maleic anhydride). Other options include N- terminal reversible bead specificity (e.g., FPCA, maleic anhydride); C-terminal specific attachment (e.g., oxazolone). An exemplary N-terminal specific resin binding (FPCA) approach is shown in FIG. 86, which utilizes click-chemistry attachment (boxed). An amine / N-terminal specific resin binding approach is shown in both FIG. 86A and FIG. 86B. It was found that adjusting for isoelectric point (e.g., reducing pH to ~6) made the process n-terminal specific (FIG. . The chemical specificity of attachment to the C- terminus was challenging. This is because the amino acids of aspartic acid (D) and glutamic acid (E) are nearly chemically identical to the C-terminus (see FIG. 87). The neighboring amide was required for distinguishing between C-terminus and D / E. This mechanism is used with oxazolone intermediate. A schematic of C-terminal specific modification using oxazolone is shown in FIG. 88.

[0321] FIG. 89 shows Raman spectra obtained for a bifunctional silane-azide linker (orange), and after binding a peptide FWRGDG (SEQ ID NO: 12) modified with an alkyne group via click chemistry (yellow).

[0322] FIG. 90 shows Raman spectra obtained for surface anchoring of a peptideATINFRRL (SEQ ID NO:6) to a metal surface. The Raman spectra is shown for a bifunctional thiol- NHS ester linker (brown), and after binding of the peptide to the surface via the linker (orange).

[0323] FIG. 91 shows Raman spectra obtained for peptide FFFRRR (SEQ ID NO: 13) attached to a photonic chip having a cichlid design as described herein. Raman spectra on the left shows baseline spectra of a clean sensor before incubation with peptides (brown), and after the attachment of the peptide (blue). The right three panels show spectra obtained from three different devices areas. The spectra shows distinct Raman features that match the bulk peptide, but with some variability potentially due to slight differences in molecular orientation / binding on the chip at the few molecule level.Improved sample handling

[0324] Automated liquid dispensers were used to aspirate samples from microwell plates and precisely dispense picoliter and nanoliter volumes into custom and commercially available picowell and microwell arrays, as well as flat and patterned substrates (see FIG. 92). Picoliter and nanoliter liquid dispensing allows for greater molecular packing density' upon sample drying (see FIG. 93 A). Automated liquid dispensing allows for site specific efficient sensor loading and greater molecular packing density through controlled, repeatable, and minimal sample volume minimizing droplet spread. This allows for the potential reduction of coffee-ring drying patterns, and for rapid droplet drying producing a droplet layering that provides on-chip sample concentration, as shown in FIG. 93B for peptide FSCFNPKCLL (SEQ ID NO: 14).

[0325] Droplet deposition into micro picowells concentrated sample through the containment of sample into a designated area (see FIGS. 94A-94B). Raman spectrum from 100 pg of dye-tagged peptide, collected in 1 second on a commercially available Vitrion picowell is shown in FIG. 94A. Raman spectrum from 12 ng of peptide, collected in 30 seconds on custom designed and fabricated microwells is shown in FIG. 94B.Improved surface attachment

[0326] It was found that surface treatments of the photonic chip further improve molecular localization, such as to localize attachment to certain materials on the chip or well (e.g., resonant features such asmetal antennae). See FIG. 95 A. Surface treatments decreased surface contamination by cleaning off surface organics prior to deposition; increased droplet contact angle further minimizing droplet spread; and increase surface anti-biofouling properties to minimize unwanted molecular adherence along microwell and picowell walls. The surface treatments can be 02 plasma, alkyl-thiols (e.g., CH3(CH2)n- SH); fluoro-alkyl-thiols (e.g., (CF3)(CF2)n(CH2)n-SH): alkyl-silanes (e.g., CH3(CH2)n-Silane); fluoro- alkyl-silanes (e.g.. (CF3)(CF2)n(CH2)n-Silane); zwitterion-functionalized molecules (e.g., SH-C11-EG4- Carboxybetaine; or 3-{[dimethyl(3-trimethoxysilyl)propyl]ammonio}propane-l-sulfonate). FIG. 94B (left) shows a microscopic image of a peptide TGYVERSPR (SEQ ID NO: 15) printed on dodecanethiol functionalized gold-coated chip. FIG. 95B (right) shows the same printed on piranha cleaned, and plasma cleaned oxide-coated gold.Single i ’x multiple molecule copies per sensor

[0327] In the case of single molecule detection, each sensing location contains only one molecule. The Raman spectra from each location (or series of spectra) is associated with a single peptide / protein sequence. Many locations and molecules can be measured and quantities of specific peptides can be determined by counting occurrences (see FIGS. 96A-96B). Peptides ASMTNMELM (SEQ ID NO: 16); KSFHFYCPL (SEQ ID NO: 17); YTICNPCVPC (SEQ ID NO: 18)

[0328] In the case of multiple molecules within a single sensing area, the Raman spectra contains signal from all the sequences. Demixing or spectral deconvolution algorithms (e.g., machine learning models) can be used to predict proportions of signal from particular sequences in the mixture. Sequences can be predicted with -80% accuracy, even when a spectra contains 20+ different sequences (see FIGS. 97A- 97B).

[0329] While preferred embodiments of the present invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. It is not intended that the invention be limited by the specific examples provided within the specification. While the invention has been described with reference to the aforementioned specification, the descriptions and illustrations of the embodiments herein are not meant to be construed in a limiting sense. Numerous variations, changes, and substitutions will now occur to those skilled in the art without departing from tire invention. Furthennore, it shall be understood that all aspects of the invention are not limited to the specific depictions, configurations or relative proportions set forth herein which depend upon a variety of conditions and variables. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in practicing the invention. It is therefore contemplated that the invention shall also cover any such alternatives, modifications, variations,or equivalents. It is intended that the following claims define the scope of the invention and that methods and structures within the scope of these claims and their equivalents be covered thereby.

Claims

CLAIMSWhat is claimed is:

1. A method of identifying a biological molecule, comprising:(a) providing a resonator comprising a nanostructure:(b) bringing said resonator in contact with said biological molecule under conditions sufficient to permit said biological molecule to couple to said nanostructure;(c) exposing said biological molecule to a first light;(d) subsequent to (c), measuring a second light emanating from said biological molecule; and(e) using at least said second light to identify said biological molecule.

2. The method of claim 1, where said biological molecule is a single biological molecule.

3. The method of claim 1 or 2, wherein said resonator comprises a first element and a second element separated by a gap, wherein said nanostructure is disposed in said gap.

4. Tire method of claim 3, wherein said gap is a nanogap.

5. The method of any one of the preceding claims, wherein said resonator is part of a chip, and wherein said method further comprises providing a mixture comprising said biological molecule to said chip.

6. The method of any one of the preceding claims, wherein said nanostructure is functionalized.

7. The method of claim 6, wherein said nanostructure is chemically functionalized.

8. Tire method of claim 7, wherein said functionalized nanostructure is functionalized with one or more moieties that reduce non-specific binding.

9. The method of claim 7 or 8. wherein said functionalized nanostructure is functionalized with one or more of the following moieties:(a) an alkyl-thiol;(b) an alkyl -oxysilane;(c) an alkyl-selenol;(d) an oxygen containing alkyl-thiol;(e) oxygen containing alkyl-oxysilanes;(f) oxygen containing alkyl-selenol;(g) zwitterionic containing thiols;(h) zwitterionic containing oxysilanes;(i) zwitterionic containing selenol;(j) cyclic containing thiols:(k) cyclic containing oxysilanes;(l) cyclic containing selenol;(m) azide containing thiols;(n) azide containing oxysilanes;(o) azide containing selenol;(p) alkyne containing thiols(q) alkyne containing oxysilanes;(r) alkyne containing selenol; or(s) molecules containing all or some of the following:(i) oxygen groups(ii) alkyl groups;(iii) a zwitterionic group.

10. The method of any one of claims 7-9, wherein said chemical functionalization comprises reducing a density of a tethering molecule in a non-specific binding reducing agent.

11. The method of any one of claims 7-10, wherein said chemical functionalization comprises backfilling into an established matrix.

12. The method of any one of claims 7-11, wherein said chemical functionalization comprises codepositing with an established matrix.

13. The method of any one of the preceding claims, wherein said biological molecule comprises a protein or peptide.

14. The method of claim 13, wherein said protein is a major histocompatibility com-plex (MHC) protein.

15. The method of claim 14, wherein said MHC protein comprises a class I peptide.

16. The method of claim 14, wherein said MHC protein comprises a class II peptide.

17. The method of claim 13, wherein said protein comprises a protein isoform.

18. The method of claim 13, wherein said protein comprises an antibody.

19. The method of any one of the preceding claims, wherein said biological molecule comprises a biologic.

20. The method of any one of claims 1-12, wherein said biological molecule comprises a nucleic acid molecule.

21. The method of claim 20, wherein said nucleic acid molecule comprises a deoxyribonucleic acid (DNA) molecule.

22. The method of claim 20, wherein said nucleic acid molecule comprises a ribonucleic acid (RNA) molecule.

23. The method of any one of claims 1-12, wherein said biological molecule comprises a sugar molecule.

24. Tire method of any one of claims 1-23, wherein said biological molecule is configured to be detected, and wherein a signal intensity associated with said biological molecule provided on said chip is at least two times higher than a signal intensity associated with a biological molecule not provided on said chip.

25. A chip for identifying a biological sample, said chip comprising a resonator comprising a nanostructure configured to attach a biological molecule of said biological sample to said nanostructure.

26. Tire chip of claim 25, where said biological molecule is a single biological molecule.

27. The chip of claim 25 or 26, wherein said resonator comprises a first element and a second element separated by a gap. wherein said nanostructure is disposed in said gap.

28. The chip of claim 27, wherein said gap is a nanogap.

29. The chip of any one of claims 25-28, wherein said resonator is part of a chip, and wherein said method further comprises providing a mixture comprising said biological molecule to said chip.

30. The chip of any one of claims 25-29, wherein said nanostructure is a functionalized nanostructure.

31. The chip of claim 30, wherein said functionalized nanostructure is chemically functionalized.

32. The chip of claim 31, wherein said functionalized nanostructure is functionalized with one or more moieties that reduce non-specific binding.

33. Tire chip of claim 31 or 32, wherein said functionalized nanostructure is functionalized with one or more of the following moieties:(a) an alkyl-thiol;(b) an alkyl -oxysilane;(c) an alkyl -sei enol;(d) an oxygen containing alkyl-thiol;(e) oxygen containing alkyl-oxysilanes;(f) oxygen containing alkyl-selenol;(g) zwitterionic containing thiols:(h) zwitterionic containing oxysilanes;(i) zwitterionic terminated selenol;(j) cyclic containing thiols;(k) cyclic containing oxysilancs;(l) cyclic containing selenol;(m) azide containing thiols:(n) azide containing oxysilanes;(o) azide containing selenol;(p) alkyne containing thiols(q) alkyne containing oxysilanes;(r) alkyne containing selenol; or(s) molecules containing all or some of the following:(i) oxygen groups;(ii) alkyl groups; or(iii) a zwitterionic group.

34. The chip of any one of claims 31-33, wherein said chemical functionalization comprises reducing a density of a tethering molecule in a non-specific binding reducing agent.

35. The chip of any one of claims 31-34, wherein said chemical functionalization comprises backfilling into an established matrix.

36. Tire chip of any one of claims 31-35, wherein said chemical functionalization comprises codepositing with an established matrix.

37. The chip of any one of claims 25-36, wherein said biological molecule comprises a protein or peptide.

38. The chip of claim 37, wherein said protein is a major histocompatibility complex (MHC) protein.

39. The chip of claim 38, wherein said MHC protein comprises a class I peptide.

40. Tire chip of claim 38, wherein said MHC protein comprises a class II peptide.

41. The chip of claim 37, wherein said protein comprises a protein isoform.

42. The chip of claim 37, wherein said protein comprises an antibody.

43. The chip of any one of claims 25-36, wherein said single biological molecule comprises a biologic.

44. The chip of any one of claims 25-36, wherein said biological molecule comprises a nucleic acid molecule.

45. The chip of claim 44, wherein said nucleic acid molecule comprises a deoxyribo-nucleic acid (DNA) molecule.

46. The chip of claim 44, wherein said nucleic acid molecule comprises a ribonucleic acid (RNA) molecule.

47. The chip of any one of claims 25-36, wherein said biological molecule comprises a sugar molecule.

48. The chip of any one of claims 25-47, wherein said biological molecule is configured to be detected, and wherein a signal intensity associated with said single biological molecule provided on said chip is at least two times high-er than a signal intensity associated with a biological molecule not provided on said chip.

49. The method of any one of claims 13-18. wherein said biological molecule is identified by cleaving an amino acid from said protein or peptide.

50. The method of claim 49, wherein said amino acid is cleaved from the N-terminus of said protein or peptide.

51. The method of claim 49 or 50, wherein cleaving said amino acid further comprises serial introduction of chemicals capable of performing said cleaving.

52. The method of claim 1, further comprising generating a synthetic spectrum.

53. The method of claim 52, wherein the synthetic spectrum is a close approximation to an empirical spectrum.

54. The method of claim 52 or 53, wherein the synthetic spectrum is developed by signal processing pipelines and machine learning models.

55. The method of any one of 52-54, wherein the synthetic spectrum is used to augment datasets, wherein the augmented data sets further comprises training a machine learning model.

56. The method of anyone of claims 52-55, wherein machine learning models are trained to identify said biological molecule based on a spectrum.

57. The method of claim 56, wherein the machine learning models are trained with any one of amino acid sequences, nucleic acid sequences, post translational modifications, secondary structures, higher-order structures, or any other chemical composition or property attributes.

58. The method of any one of claims 52-57. wherein a spectrum is generated through chemical, electrical, or physical perturbations.

59. The method of claim 58, wherein perturbations are generated from heating, charge or electric field perturbations, pH perturbations, or chemical binding and cleavages.

60. Tire method of any prior claim, further comprising estimating a three-dimensional structure of the biological molecule61. The method of claim 60, wherein a machine learning model estimates the three-dimensional structure of the biological molecule.

62. The method of claim 61, wherein the machine learning model estimates the three-dimensional structure based on post translational modifications.

63. Tire method of claim 61, wherein the machine learning model estimates tire three-dimensional structure based on chemical bond energies.

64. The method of claim 62 or 63, wherein the post translational modification and / or the chemical bond energies generate at least 10% higher specificity and / or confidence values than existing structure prediction algorithms (e.g., Al-phaFold, AlphaFold2, PepFold).

65. Tire method of any prior claim, wherein the machine learning model predicts a density or a number of molecules coupled to said nanostructure.

66. The method of claim 65, wherein predicting comprises generating probabilities for the number of molecules coupled to said nanostructure.

67. The method of claim 65 or 66, wherein said density or said number of molecules corresponds to a concentration of said molecules.

68. Tire method of any prior claim, wherein the machine learning model predicts type or types of molecules bound to the nanostructure.

69. The method of claim 68, further comprising predicting the proportion of the types of molecules bound to the nanostructure.

70. The method of any prior claim, wherein the machine learning model generates a quantification estimate for each type of molecule analyzed over an entire array.

71. The method of claim 1, further comprising generating an optical readout.

72. The method of claim 71, wherein the optical readout is based on a vibrational fingerprint of the biological molecule.

73. The method of claim 71, wherein the optical readout is based on a signal intensity of the biological molecule.

74. Tire method of claim 72 or 73, wherein the resonator is configured to enhance the vibrational fingerprint and / or signal intensity of the biological molecule.

75. The method of claim 74, wherein the resonator enhances the vibrational finger-print and / or signal intensity of a biological molecule coupled to the resonator at least two times as compared to a biological molecule not interacting with the resonator.

76. The method of claim 74, wherein the resonator enhances the vibrational finger-print and / or signal intensity of a biological molecule interacting with the resonator at least 102 times as compared to a biological molecule not inter-acting with the resonator.

77. The method of claim 74, wherein the resonator enhances the vibrational finger-print and / or signal intensity of a biological molecule interacting with the resonator at least 103 times as compared to a biological molecule not inter-acting with the resonator.

78. The method of claim 74, wherein the resonator enhances the vibrational finger-print and / or signal intensity of a biological molecule interacting with the resonator at least 105 times as compared to a biological molecule not inter-acting with the resonator.

79. The method of any prior claim, wherein the nanostructure is metallic.

80. The method of any prior claim, wherein the resonator is an optical resonator.

81. Tire method of any prior claim wherein the resonator is a dielectric resonator.

82. Tire method of any prior claim, wherein the resonator is an optical dielectric resonator.

83. The method of any prior claim, wherein the resonator is a semi conducting resonator.

84. The method of any prior claim, wherein the second light is a different frequency than the first light.

85. The method of any prior claim, wherein said biological molecule comprises no more than ten biological molecules.

86. Tire method of claim 13, wherein said protein comprises a human leukocyte anti-gen (HLA).

87. The method of claim 86, wherein said HLA a class I peptide.

88. The method of claim 86, wherein said HLA a class II peptide.

89. The chip of claim 37, wherein said protein comprises a human leukocyte antigen (HLA).

90. The chip of claim 89, wherein said HLA a class I peptide.

91. Tire chip of claim 89, wherein said HLA a class II peptide.

92. The chip of any claim, wherein said biological molecule comprises no more than ten biological molecules.

93. The method or chip of prior claim, wherein single molecule vibrational spectroscopy is for fingerprinting and / or sequencing of:(a) major histocompatibility complex (MHC) class I and / or II peptides;(b) antibodies and / or antibody fragments;(c) biologies, wherein said biologies include post-translational modifications and / or secondary / tertiary structure:(d) protein isoforms and / or protein fragments;(e) DNA and / or complementary DNA (cDNA); or(f) any combination thereof94. Tire method of any prior claim, wherein the identifying of a molecule is performed de novo.

95. The method of claim 52 or 5533, wherein the synthetic spectrum is developed by using ab-initio calculations.

96. The method of any one of claims 52-54. further comprising generating the synthetic spectrum using ab-initio calculations.

97. The method of claim 52or 53, wherein the synthetic spectrum is developed by using molecular dynamics simulations.

98. The method of any one of claims 52-54, further comprising generating the synthetic spectrum using molecular dynamics simulations.

99. Tire method of claim 74, wherein the resonator enhances the vibrational finger-print and / or signal intensity of a biological molecule interacting with the resonator at least 1010 times as compared to a biological molecule not interacting with the resonator.

100. The method of claim 74, wherein the resonator enhances the vibrational finger-print and / or signal intensity of a biological molecule interacting with the resonator at least 1015 times as compared to a biological molecule not interacting with the resonator.

101. Tire method of claim 74, wherein the resonator enhances the vibrational finger-print and / or signal intensity of a biological molecule interacting with the resonator at least 1020 times as compared to a biological molecule not interacting with the resonator.

102. The method of claim 1 or 2. wherein said resonator comprises a first element and a second element separated by a gap, wherein said nanostructure is disposed atop said gap.

103. The chip of claim 25 or 26, wherein said resonator comprises a first element and a second element separated by a gap, wherein said nanostructure is disposed atop said gap.

104. The method of claim 102 or the chip of claim 103, wherein said nanostructure disposed atop said gap is disposed on a top face of a resonator element.

105. A photonic chip, comprising a substrate layer and an array of one or more photonic unit cells, wherein the array of the one or more photonic unit cells is positioned above and adjacent to the substrate layer, and each of the photonic unit cells comprises a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap, and wherein(a) the primary photonic pillar is an ellipse or polygon comprising a dielectric base layer, dielectric spacer layer, and one or more photonic antennae, wherein(i) the dielectric base layer of the primary photonic pillar is positioned above and adjacent to the substrate layer;(ii) the dielectric spacer layer of the primary photonic pillar is positioned above and adjacent to the dielectric base layer of the primary photonic pillar; and(iii) the one or more photonic antennae are positioned above and adjacent to the dielectric spacer layer of the primary photonic pillar, and(b) each of the one or more secondary photonic pillars is an ellipse or polygon comprising a dielectric base layer, wherein(i) the dielectric base layer of the one or more secondary photonic pillars is positioned above and adjacent to the substrate layer.

106. The photonic chip of claim 105, having an arrangement of components as shown in the schematic of FIG. 52B.

107. A photonic chip, comprising a substrate layer and an array of one or more photonic unit cells, wherein the array of the one or more photonic unit cells is positioned above and adjacent to the substrate layer, and each of the photonic unit cells comprises a primary’ photonic pillar and one or more secondary’ photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap, and wherein(a) the primary photonic pillar is an ellipse or polygon comprising a dielectric base layer, dielectric spacer layer, and one or more photonic antennae, wherein(i) the dielectric base layer of the primary photonic pillar is positioned above and adjacent to the substrate layer;(ii) the dielectric spacer layer of the primary photonic pillar is positioned above and adjacent to the dielectric base layer of the primary photonic pillar; and(iii) the one or more photonic antennae are positioned above and adjacent to the dielectric spacer layer of the primary photonic pillar,(b) each of the one or more secondary photonic pillars is an ellipse or polygon comprising a dielectric base layer and dielectric spacer layer, wherein(i) the dielectric base layer of the one or more secondary photonic pillars is positioned above and adjacent to the substrate layer; and(ii) the dielectric spacer layer of the one or more secondary photonic pillars is positioned above and adjacent to the dielectric base layer of the one or more secondary' photonic pillars, and the substrate layer comprises a silicon layer and an oxide layer, wherein the oxide layer is above and adjacent to the silicon layer.

108. The photonic chip of claim 107, having an arrangement of components as shown in the schematic of FIG. 68A.

109. A photonic chip, comprising a substrate layer and an array of one or more photonic unit cells, wherein the array of the one or more photonic unit cells is positioned above and adjacent to the substrate layer, and each of the photonic unit cells comprises a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap, and wherein(a) the primary photonic pillar is an ellipse or polygon comprising a dielectric base layer, dielectric spacer layer, and one or more photonic antemiae, wherein(i) the dielectric base layer of the primary photonic pillar is positioned above and adjacent to the substrate layer;(ii) the dielectric spacer layer of the primary photonic pillar is positioned above and adjacent to the dielectric base layer of the primary photonic pillar; and(iii) the one or more photonic antennae are positioned above and adjacent to the dielectric spacer layer of the primary photonic pillar,(b) each of the one or more secondary’ photonic pillars is an ellipse or polygon comprising a dielectric base layer and dielectric spacer layer, wherein(i) the dielectric base layer of the one or more secondary photonic pillars is positioned above and adjacent to the substrate layer; and(ii) the dielectric spacer layer of the one or more secondary photonic pillars is positioned above and adjacent to the dielectric base layer of the one or more secondary photonic pillars, and the substrate layer comprises a metal layer and an oxide layer, wherein the oxide layer is above and adjacent to the metal layer.

110. Tire photonic chip of claim 109, having an arrangement of components as shown in the schematic of FIG. 68B.

111. A photonic chip, comprising a substrate layer, a dielectric fill layer, and an array of one or more photonic unit cells, yvherein the array of the one or more photonic unit cells is positioned above and adjacent to the substrate layer, each of the photonic unit cells comprises a primary photonic pillar and one or more secondaryphotonic pillars, wherein each of the primary photonic pillar and one or more secondary' photonic pillars are separated by a gap, and wherein(a) the primary photonic pillar is an ellipse or polygon comprising a dielectric base layer, a portion of the dielectric fill layer, and one or more photonic antennae, wherein(i) the dielectric base layer of the primary photonic pillar is positioned above and adjacent to the substrate layer;(ii) the portion of the dielectric fill layer of the primary photonic pillar is positioned above and adjacent to the dielectric base layer of the primary photonic pillar, and(iii) the one or more photonic antennae are positioned above and adjacent to the portion of the dielectric fill layer of the primary7photonic pillar, and(b) each of the one or more secondary7photonic pillars is an ellipse or polygon comprising a dielectric base layer and a portion of the dielectric fill layer, wherein(i) the dielectric base layer of the one or more secondary photonic pillars is positioned above and adjacent to the substrate layer; and(ii) the portion of the dielectric fill layer of the secondary' photonic pillar is positioned above and adjacent to the dielectric base layer of the secondary photonic pillar, and wherein the dielectric fill layer covers the surface of and the gap between each of the primary and secondary photonic pillars and is below the one or more photonic antennae of the primary photonic pillar.

112. The photonic chip of claim 111, having an arrangement of components as shown in the schematic of FIG. 68C.

113. A photonic chip, comprising a substrate layer, a dielectric fill layer, an array of one or more photonic unit cells, and a passivation layer, wherein the array of the one or more photonic unit cells is positioned above and adjacent to the substrate layer, each of the photonic unit cells comprises a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap, and wherein(a) the primary photonic pillar is an ellipse or polygon comprising a dielectric base layer, a portion of the dielectric fill layer, one or more photonic antennae, and a portion of the passivation layer, wherein(i) the dielectric base layer of the primary photonic pillar is positioned above and adjacent to the substrate layer;(ii) the portion of the dielectric fill layer of the primary photonic pillar is positioned above and adjacent to the dielectric base layer of the primary photonic pillar;(iii) the one or more photonic antennae are positioned above and adjacent to the portion of the dielectric fill layer of the primary photonic pillar, and(iv) the portion of the passivation layer of the primary photonic pillar is positioned above and adjacent to the one or more photonic antennae and the portion of the fill layer of the primary photonic pillar; and(b) each of the one or more secondary photonic pillars is an ellipse or polygon comprising a dielectric base layer, a portion of the dielectric fill layer, and a portion of the passivation layer, wherein(i) the dielectric base layer of the one or more secondary photonic pillars is positioned above and adjacent to the substrate layer;(ii) the portion of the dielectric fill layer of the secondary photonic pillar is positioned above and adjacent to the dielectric base layer of the secondary photonic pillar;(iii) the portion of the passivation layer of the secondary photonic pillar is positioned above and adjacent to the portion of the dielectric fill layer of the secondary photonic pillar, and wherein the dielectric fill layer covers the surface of and the gap between each of the primary and secondary photonic pillars and is below the one or more photonic antennae of the primary photonic pillar, and the passivation layer covers the surface of the dielectric fill layer, optionally wherein the passivation layer comprises a gap in coverage at the one or more photonic antennae, wherein the gap in coverage is an absence of passivation layer.1 14. The photonic chip of claim 1 13, having an arrangement of components as shown in the schematic of FIG. 68D.

115. A photonic chip, comprising a substrate layer, a dielectric fill layer, and an array of one or more photonic unit cells, wherein the array of the one or more photonic unit cells is positioned above and adjacent to the substrate layer, each of the photonic unit cells comprises a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap, and wherein(a) the primary photonic pillar is an ellipse or polygon comprising a dielectric base layer, a portion of the dielectric fill layer, and one or more photonic antennas, wherein(i) the dielectric base layer of the primary photonic pillar is positioned above and adjacent to the substrate layer;(ii) the portion of the dielectric fill layer of the primary photonic pillar is positioned above and adjacent to the dielectric base layer of the primary photonic pillar, and(iii) the one or more photonic antennae is positioned above and adjacent to the portion of the dielectric fill layer of the primary photonic pillar, and(b) each of the one or more secondary photonic pillars is an ellipse or polygon comprising a dielectric base layer and a portion of the dielectric fill layer, wherein(i) the dielectric base layer of the one or more secondary photonic pillars is positioned above and adjacent to the substrate layer; and(ii) the portion of the dielectric fill layer of the secondary photonic pillar is positioned above and adjacent to the dielectric base layer of the secondary photonic pillar, and wherein the dielectric fill layer covers the surface of and the gap between each of the primary and secondary photonic pillars and is below the one or more photonic antennae of the primary photonic pillar, and the substrate layer comprises a metal layer or dielectric mirror layer and dielectric spacer layer, wherein the dielectric spacer layer is above and adjacent to the metal layer or dielectric mirror layer.

116. The photonic chip of claim 115. having an arrangement of components as shown in the schematic of FIG. 69A.

117. A photonic chip, comprising a substrate layer, a dielectric fill layer, and an array of one or more photonic unit cells, wherein the array of the one or more photonic unit cells is positioned above and adjacent to the substrate layer, each of the photonic unit cells comprises a primary photonic pillar and one or more secondary' photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap, and wherein(a) the primary photonic pillar is an ellipse or polygon comprising a dielectric base layer, a portion of the dielectric fill layer, and one or more photonic antennae, wherein(i) the dielectric base layer of the primary photonic pillar is positioned above and adjacent to the substrate layer;(ii) the portion of the dielectric fill layer of the primary photonic pillar is positioned above and adjacent to the dielectric base layer of the primary photonic pillar, and(iii) the one or more photonic antennae is positioned above and adjacent to the portion of the dielectric fill layer of the primary photonic pillar, and(b) each of the one or more secondary photonic pillars is an ellipse or polygon comprising a dielectric base layer and a portion of the dielectric fill layer, wherein(i) the dielectric base layer of the one or more secondary' photonic pillars is positioned above and adjacent to the substrate layer; and(ii) the portion of the dielectric fill layer of the secondary photonic pillar is positioned above and adjacent to the dielectric base layer of the secondary photonic pillar, and wherein the dielectric fill layer covers the surface of and the gap between each of the primary and secondary photonic pillars and is below the one or more photonic antennae of the primary photonic pillar, and the substrate layer comprises a silicon or glass wafer layer, a metal layer or dielectric mirror layer, and a dielectric spacer layer, wherein the metal layer or dielectric mirror layer is above and adjacent to the silicon or glass wafer layer, and the dielectric spacer layer is above and adjacent to the metal layer or dielectric mirror layer.

118. Tire photonic chip of claim 117, having an arrangement of components as shown in the schematic of FIG. 69D.

119. A photonic chip, comprising a substrate layer, a dielectric fill layer, and an array of one or more photonic unit cells, wherein the array of the one or more photonic unit cells is positioned above and adjacent to the substrate layer, each of the photonic unit cells comprises a primary photonic pillar and one or more secondary photonic pillars, wherein each of the primary photonic pillar and one or more secondary photonic pillars are separated by a gap, and wherein(a) the primary photonic pillar is an ellipse or polygon comprising a dielectric base layer, a portion of the dielectric fill layer, and one or more photonic antennae, wherein(i) the dielectric base layer of the primary photonic pillar is positioned above and adjacent to the substrate layer;(ii) the portion of the dielectric fill layer of the primary photonic pillar is positioned above and adjacent to the dielectric base layer of the primary photonic pillar, and(iii) the one or more photonic antennae is positioned above and adjacent to the portion of the dielectric fdl layer of the primary photonic pillar, and(b) each of the one or more secondary' photonic pillars is an ellipse or polygon comprising a dielectric base layer and a portion of the dielectric fdl layer, wherein(i) the dielectric base layer of the one or more secondary photonic pillars is positioned above and adjacent to the substrate layer; and(ii) the portion of the dielectric fdl layer of the secondary' photonic pillar is positioned above and adjacent to the dielectric base layer of the secondary photonic pillar, and wherein the dielectric fdl layer covers the surface of and the gap between each of the primary and secondary photonic pillars and is below the one or more photonic antennae of the primary photonic pillar, and the substrate layer comprises a silicon or glass wafer layer, alternating layers of a silicon layer and an oxide layer, and a dielectric spacer layer, wherein the alternating layers of a silicon layer and an oxide layer are above and adjacent to the silicon or glass wafer layer, and the dielectric spacer layer is above and adjacent to the alternating layers of a silicon layer and an oxide layer.

120. The photonic chip of claim 1 19, having an arrangement of components as shown in the schematic of FIG. 69E.

121. The photonic chip of any one of claims 105-120, wherein the dielectric base layer is silicon or a silicon-based material.

122. Tire photonic chip of any one of claims 105-120, wherein the one or more photonic antennae comprises or consists of a metal or metal composition.

123. The photonic chip of claim 121, wherein the photonic antenna comprises or consists of gold, silver, platinum, copper, aluminum, or titanium nitride.

124. The photonic chip of any one of claims 105-123, wherein the one or more photonic antenna is a single resonant structure or double resonant structure.

125. The photonic chip of claim 124. wherein the single resonant structure comprises a shape selected from a triangle, a circle, an ellipse, a trapezoid, a rectangle, a square, or another shape having single resonance.

126. A method of manufacturing a photonic chip, the method comprisingI l l(a) providing a starting substrate:(b) depositing a mirror layer on top of the starting substrate, wherein the mirror layer comprises one or more alternating layers of a silicon layer and an oxide layer, or a mirror layer;(c) depositing a dielectric spacer on top of the mirror layer:(d) depositing a dielectric device layer on top of the dielectric spacer layer, optionally wherein the dielectric device layer is a silicon device layer;(e) depositing a first mask layer on top of the dielectric device layer;(f) depositing a first photoresist layer on top of the first mask layer;(g) performing photolithography patterning on the first photoresist layer;(h) developing the first photoresist layer;(i) etching the first mask layer and the dielectric device layer;(j) removing the first mask;(k) depositing a dielectric fill layer on top of the exposed dielectric spacer and on the exposed dielectric device layer;(l) depositing layer of material for one or more photonic antennae on top of the dielectric fill layer;(m) depositing a second mask layer on top of the layer of material for the one or more photonic antennae;(n) depositing a second photoresist layer on top of the second mask layer;(o) performing photolithography patterning on the second photoresist layer;(p) developing the second photoresist layer;(q) etching the second mask layer and the layer of material for one or more photonic antennae: and(r) removing the second mask layer, thereby producing the photonic chip.

127. The method of claim 126. wherein tire dielectric spacer layer is an oxide, optionally wherein the oxide is a silicon dioxide.

128. The method of claim 126 or 127, wherein the dielectric device layer is a silicon device layer.

129. The method of claim 128, wherein the silicon device layer comprises crystalline silicon, amorphous silicon, or silicon nitride.

130. The method of claim 128, wherein the silicon device layer comprises crystalline silicon, amorphous silicon, or silicon nitride.

131. The method of any one of claims 126-130, wherein the exposed dielectric device layer of step (k) comprises a dielectric layer of one or more photonic pillar.

132. Tire method of any one of claims 126-130, wherein the one or more photonic antennae comprises a metal or metal composition.

133. A method of determining a cellular response of a cell to a stimulus, comprising(a) providing a cell and a photonic chip, wherein the chip comprises a resonator;(b) attaching the cell to the photonic chip at or adjacent to the resonator;(c) exposing the cell to a first light;(d) measuring a second light emanating from the cell; and(e) using at least the second light to identify the cellular response of the cell, wherein the stimulus is applied to the cell before attachment to the photonic chip, or after attachment to the photonic chip.

134. The method of claim 133. wherein the second light is a vibrational spectral signature.

135. Tire method of claim 134, wherein the vibrational spectral signature is a Raman spectral signature or an infrared (IR) spectral signature.

136. The method of any one of claims 133-135, wherein the method is non-destractive to the cell.

137. The method of any one of claims 133-136, wherein the method is label free.

138. The method of any one of claims 133-137, wherein machine learning models are trained to identify said cellular response of a cell based on a spectrum.

139. The method of any one of claims 133-138, wherein the cell is a mammalian cell, optionally a human cell.

140. The method of claim 139, wherein the cell is a human T cell.

141. The method of claim 140, wherein machine learning models are trained to identify the cellular response of a T cell or state of a T cell based on Raman fingerprint interpretability determined by machine learning methods.

142. The method of claim 141, wherein T cell differentiation or T cell sub-type is determined.

143. Tire method of claim 142, wherein the T cell differentiation is selected from CD4+, CD8+, thymic Treg, mucosal associated invariant T cell (MAIT). invariant natural killer T cell (iNKT), or induced Treg (iTreg).

144. Tire method of claim 142, wherein the T cell sub-type is selected from Thl, Th2, Tel, Tc2, or Trm.

145. A macromolecule, comprising(a) a Raman reporter molecule; and(b) a barcoded peptide-MHC protein complex, optionally wherein the barcode is a DNA barcode.

146. The macromolecule of claim 145, wherein(a) the Raman reporter molecule is an aromatic molecule having a unique Raman fingerprint; or(b) the barcoded peptide-MHC protein complex is a pMHC-dextramer.

147. The macromolecule of claim 145 or 146, wherein the barcoded peptide-MHC protein complex comprises a fluorophore.

148. The macromolecule of any one of claims 145-147, for use in vibrational spectroscopic determination of(a) a cellular response of a cell to a stimulus;(b) a cellular state of a cell;(c) a cellular type of a cell; or(d) a cellular binding activity of a cell.

149. Tire Raman macromolecule for use according to claim 148, wherein machine learning models are trained to identify(a) the cellular response of the cell to the stimulus;(b) the cellular state of the cell;(c) the cellular type of the cell; or(d) the cellular binding activity of the cell based on Raman fingerprint interpretability determined by machine learning methods.

150. The Raman macromolecule for use according to claim 149, wherein machine learning models are trained to demix the barcode identities when multiple barcodes are simultaneously present.

151. A method of producing a Raman macromolecule, the method comprising(a) providing each of(i) a Raman reporter molecule; and(ii) a barcoded peptide-MHC protein complex, and(b) attaching the particle and Raman reporter molecule of step (b) to the barcoded peptide- MHC protein complex.

152. The method of claim 150, wherein(a) the particle is a microbead or nanoparticle, optionally wherein the particle is porous;(b) the Raman reporter molecule is an aromatic molecule having a unique Raman fingerprint; or(c) the barcoded peptide-MHC protein complex is a pMHC-dextramer.

153. A Raman macromolecule, comprising(a) a particle;(b) a Raman reporter molecule; and(c) a barcoded peptide-MHC protein complex, optionally wherein the barcode is a DNA barcode.

154. The Raman macromolecule of claim 153, wherein(a) the particle is a microbead or nanoparticle, optionally wherein tire particle is porous;(b) the Raman reporter molecule is an aromatic molecule having a unique Raman fingerprint; or(c) the barcoded peptide-MHC protein complex is a pMHC-dextramer.

155. The Raman macromolecule of claim 153 or 154, wherein(a) the particle is attached to the Raman reporter molecule noncovalently, optionally wherein the noncovalent attachment is an intercalation of the Raman reporter molecule into the particle;(b) the particle is attached to the barcoded peptide-MHC protein complex with a cleavable linkage, optionally wherein the cleavable linkage is a disulfide linkage.

156. The Raman macromolecule of any one of claims 153-155, wherein the barcoded peptide-MHC protein complex comprises a fluorophore.

157. The Raman macromolecule of any one of claims 153-156, for use in vibrational spectroscopic determination of(a) a cellular response of a cell to a stimulus;(b) a cellular state of a cell;(c) a cellular type of a cell; or(d) a cellular binding activity of a cell.

158. The Raman macromolecule for use according to claim 157, wherein machine learning models are trained to identify(a) the cellular response of the cell to the stimulus;(b) the cellular state of the cell;(c) the cellular type of the cell; or(d) the cellular binding activity of the cell based on Raman fingerprint interpretability determined by machine learning methods.

159. The Raman macromolecule for use according to claim 158, wherein machine learning models are trained to demix the barcode identities when multiple barcodes are simultaneously present.

160. A method of producing a Raman macromolecule, the method comprising(a) providing each of(i) a particle;(ii) a Raman reporter molecule: and(iii) a barcoded peptide-MHC protein complex;(b) attaching the particle to the Raman reporter molecule; and(c) attaching the particle and Raman reporter molecule of step (b) to the barcoded peptide- MHC protein complex.

161. The method of claim 160, wherein(a) the particle is a microbead or nanoparticle, optionally wherein the particle is porous;(b) the Raman reporter molecule is an aromatic molecule having a unique Raman fingerprint; or(c) the barcoded peptide-MHC protein complex is a pMHC-dextramer.

162. The method of claim 160 or 161, wherein the(a) the attachment of the particle to the Raman reporter molecule is noncovalent, optionally wherein the noncovalent attachment is an intercalation of the Raman reporter molecule to the particle; or(b) the attachment of the particle to the barcoded peptide-MHC protein complex is with a cleavable linkage, optionally wherein the cleavable linkage is a disulfide linkage.

163. A peptide-pulsed antigen presenting cell (APC), comprising(a) an APC;(b) at least one antigen, wherein each antigen is an immunogenic peptide; and(c) a Raman reporter molecule, wherein the Raman reporter is attached to the APC or to the antigen, and wherein the APC is pulsed with the at least one antigen.

164. The peptide-pulsed APC of claim 163, wherein the Raman reporter molecule is an aromatic molecule having a unique Raman fingerprint.

165. Tire peptide-pulsed APC of claim 163 or 164, wherein the Raman reporter is attached to the APC using a protein, peptide, lipid, or glycan.

166. The peptide-pulsed APC of any one of claims 163-165, for use in vibrational spectroscopic determination of(a) a cellular response of an APC after stimulation with an antigen;(b) a cellular state of an APC after stimulation with an antigen;(c) a cellular type of an APC after stimulation with an antigen; or(d) a cellular binding activity of APC after stimulation with an antigen.

167. The peptide-pulsed APC for use according to claim 166, wherein machine learning models are trained to identify(a) the cellular response of the APC after stimulation with the antigen;(b) the cellular state of the APC after stimulation with the antigen;(c) the cellular type of the APC after stimulation with the antigen; or(d) the cellular binding activity of the APC after stimulation with the, based on Raman fingerprint interpretability determined by machine learning methods.

168. The Raman macromolecule for use according to claim 167, wherein machine learning models are trained to demix the barcode identities when multiple barcodes are simultaneously present.

169. A method for determining an activated or unactivated state of CD8+ T cells, the method comprising(a) contacting CD8+ T cells with a Raman macromolecule comprising a particle, and a Raman reporter molecule or a barcoded peptide-MHC protein complex, thereby producing a Raman macromolecule / T cell complex;(b) loading the Raman macromolecule / T cell complex into an array for vibrational spectra determination;(c) determining the vibrational spectra for the Raman macromolecule / T cell complex at one or more timepoints;(d) immunostaining the T cells of the Raman macromolecule / T cell complex with an antiCD 137 binding molecule, and optionally staining the T cells of the Raman macromolecule / T cell complex with propidium iodide, and further optionally imaging the T cells of the Raman macromolecule / T cell complex using fluorescent microscopy;(e) performing fluorescence-activated cell sorting (FACS) on the T cells of the Raman macromolecule / T cell complex based on anti-CD137 binding to sort for activated and unactivated CD8+ T cells, and optionally confirming results from the fluorescent microscopy imaging;(f) perfonning single-cell RNA sequencing (scRNA-seq) to determine an activated or unactivated state of the T cells.

170. Tire method of claim 169, wherein machine learning models are trained to identify(a) an activated or unactivated state of the T cells;(b) a cellular response of the T cells;(c) a cellular state of the T cells;(d) a cellular type of the T cells; or(e) a cellular binding activity of the T cells; based on Raman fingerprint interpretability determined by machine learning methods.

171. The method according to claim 170. wherein the machine learning models are trained to demix the barcode identities when multiple barcodes are simultaneously present.

172. A method for sequencing activated and exhausted CD8+ T cells, tire method comprising(a) contacting CD8+ T cells with a Raman macromolecule comprising a particle, and a Raman reporter molecule or a barcoded peptide-MHC protein complex, thereby producing a Raman macromolecule / T cell complex, wherein the Raman reporter molecule or the barcoded peptide-MHC protein complex is attached to the particle using a cleavable linkage, optionally wherein the cleavable linkage is a disulfide linkage;(b) loading the Raman macromolecule / T cell complex into an array for vibrational spectra determination; and(c) determining the vibrational spectra for the Raman macromolecule / T cell complex at one or more timepoints;(d) immunostaining the T cells of the Raman macromolecule / T cell complex with anti- CD137, anti-PD-1, anti-LAG3, and / or anti-CTLA4 binding molecules, and optionally staining the T cells of the Raman macromolecule / T cell complex with propidium iodide, and further optionally imaging the T cells of the Raman macromolecule / T cell complex using fluorescent microscopy;(e) performing fluorescence-activated cell sorting (FACS) on the T cells of the Raman macromolecule / T cell complex based on anti-CD137, anti-PD-1, anti-LAG3, and / or anti- CTLA4 binding to sort for antigen-bound active and exhausted CD8+ T cells, and optionally confirming results from the fluorescent microscopy imaging;(f) cleaving the linkage between the particle and the Raman reporter molecule or barcoded peptide-MHC protein complex of the Raman macromolecule / T cell complex and isolating activated and exhausted antigen-specific T cells;(g) amplifying the TCR sequences from the activated and exhausted antigen-specific T cells; and(h) sequencing the TCR sequences from the activated and exhausted antigen-specific T cells.

173. Hie method of claim 172, wherein machine learning models are trained to identify(a) an activated and / or exhausted state of the T cells;(b) a cellular response of the T cells;(c) a cellular state of the T cells;(d) a cellular type of the T cells; or(e) a cellular binding activity of the T cells; based on Raman fingerprint interpretability detennined by machine learning methods.

174. The method according to claim 173, wherein the machine learning models are trained to demix the barcode identities when multiple barcodes are simultaneously present.

175. The method of any one of claims 169-174, wherein the method comprises prior to step (a) the steps of(i) activating peripheral blood mononuclear cells with T cell activating peptides; and(ii) negatively selecting for CD8+ T cells.

176. The method of any one of claims 169-175, when used in a high throughput and multiplexed screening of T cells and antigens.

177. The method of any one of claims 169-175, when used in a high throughput and multiplexed screening of primary T cell activation from T cells, or peptide-pulsed antigen presenting cells.

178. A nanowell flow cell, comprising a means for continuous tracking of individual T cells for at time points before, during, and after the introduction of pMHC multimers.

179. A nanowell flow cell, comprising a means for recovering T cells for sequencing after vibrational spectra determination.

180. The method of any of the preceding claims, further comprising(a) pre-processing the sample for vibrational spectroscopy, optionally wherein the preprocessing comprises chemical modification of the analyte;(b) fractioning the sample for vibrational spectroscopy;(c) spotting or printing the sample for vibrational spectroscopy;(d) attaching the sample for vibrational spectroscopy to a photonic chip;(e) obtaining a vibrational spectra for the photonic chip (e.g.. Raman spectroscopy); or(f) analyzing the vibrational spectra Data, optionally wherein the molecular species and of the sample and quantity of the sample is determined.

181. The method of claim 180. wherein tire sample pre-processing comprises separating the sample into different fractions based on(a) solubility;(b) affinity pulldown; or(c) a resin based approach.

182. The method of claim 180, wherein the sample fractioning comprises separating the sample into one or more fractions based on physical or chemical properties of the analyte, wherein the fractionization is performed using(a) liquid chromatography;(b) electrophoretic separation;(c) surface coated microparticles;(d) or a combination thereof optionally wherein the physical or chemical properties are selected from solubility, charge, amino acid composition, side chain group chemistry, post-translational modifications, or size.

183. The method of any of the preceding claims, further comprising chemically modifying the analyte at the N-terminus using benzaldehyde, 2-pyridinecarboxaldehyde, or a compound or resin comprising benzaldehyde or 2-pyridinecarboxaldehyde.

184. The method of any of the preceding claims, further comprising chemically modify ing the analyte at the N-terminus using maleic anhydride, or a compound or resin comprising maleic anhydride.

185. Tire method of any of the preceding claims, further comprising chemically modifying the analyte at the C-temiinus using oxazolone formation at C-temiinus, optionally by treatment with acetic anhydride and fonnic acid, optionally further comprising contact with an activator, optionally wherein the activator is selected from pentafluorophenol, hydroxybenzotriazole (HOBt), hexafluorophosphate benzotriazole tetramethyl uronium (HBTU), O-(lH-6-chlorobenzotriazole-l-yl)-l, 1,3,3- tctramcthyluronium hcxafluorophosphatc (HCTU), or hcxafluorophosphatc azabcnzotriazolc tetramethyl uronium (HATU); optionally further comprising derivatization with an amine -containing compound, optionally wherein the amine-containing compound is an alkyne-amine.

186. Tire method of any of the preceding claims, further comprising modify ing a solid surface or substrate surface for covalent anchoring of an analyte with functional monolayers comprising a surface reactive group linked to a peptide reactive group, optionally wherein the surface reactive group is selected from(a) a silane or silane group, optionally selected from triethoxy silane, tri methoxy silane, or chloro-silane;(b) a thiol based group;(c) a selenol-based group, and the peptide reactive group is selected from an NHS-ester, epoxide, carbodiimide, maleimide, or maleic anhydride.

187. The method of any of the preceding claims, further comprising modifying a solid surface or substrate surface for covalent anchoring of an alkyne modified peptide molecule, comprising(a) modifying the solid surface or substrate surface with one or more functional monolayers comprising a surface reactive group linked to an azide, optionally wherein the surface reactive group is a silane or silane group, thiol, or selenol; and(b) attaching an alkyne modified peptide to the azide modified surface, optionally via treatment with copper sulfate and sodium ascorbate.

188. The method of any of the preceding claims, further comprising treating the solid surface or substrate surface to improve sample surface concentration, optionally wherein the treating comprises(a) oxygen plasma treatment(b) alkyl-thiol monolayer deposition;(c) fluoro-alkyl-thiol monolayer deposition;(d) alkyl-silane monolayer deposition;(e) aluoro-alkyl-silane monolayer deposition; or(f) zwitterion-functional monolayer deposition, optionally selected from (sh-cl l-eg4- carboxybetaine or 3-{ [dimethyl(3-trimethoxysilyl)propyl]ammonio}propane- 1 - sulfonate).

189. Any chip, array, unit cell, resonator, or nanostructure as described herein, for use in any method as described herein.

190. Any method as described herein, comprising implementation of any chip or machine learning as a described herein.

191. The method or chip of any prior claim, wherein the nanostructure comprises a metal.

192. The method or chip of claim 191, wherein the metal comprises gold, platinum, palladium, copper, aluminum, titanium, chromium, or other noble metals, coinage metals, and any combination thereof193. The method or chip of any prior claim, wherein signal processing pipelines and machine learning models are developed to synthesize spectra that closely approximate empirical spectra.

194. Tire method or chip of any prior claim, wherein machine learning models are trained and used for spectra-to-molecule identification.

195. The method or chip of any prior claim, wherein each input data point is composed of a set or series of spectra generated through chemical, electrical, or physical perturbations.

196. The method or chip of any prior claim, wherein a 3 -dimensional structure of the biological molecule is estimated using a model with input comprising PTMs and chemical bond energies.

197. Tire method or chip of any prior claim, wherein a trained model would predict a density or number of molecules coupled to the resonator.

198. The method or chip of any prior claim, wherein a trained model predicts a mixture or types of biological molecules bound to the chip.

199. The method or chip of any prior claim, wherein machine learning models are trained to provide quantification estimates for each analyte type present in an assayed sample over an entire array.

200. All that is described, drawn, or demonstrated herein.

Images