Systems and methods for spectroscopic determination of a chemometric model from sample scans

Abstract

A computer-implemented method is provided. The method includes obtaining a chemometric model for one or more parameter levels associated with composition training data. The composition training data includes data representative of one or more components of a vesicle. Receiving Raman spectral data representative of at least one Raman spectrum associated with one or more components of a sample of a vesicle. Communicating the Raman spectral data representative of the at least one Raman spectrum associated with the one or more components of the sample of the vesicle into the chemometric model. Determining a level of one or more parameters of the one or more components of the sample of the vesicle based on the chemometric model.

Description

Technical Field

[0001] This disclosure generally relates to systems and methods for performing spectroscopic analysis techniques, such as Raman spectroscopy. In particular, systems and methods are disclosed for determining stoichiometric models for one or more levels of one or more parameters from one or more components of a vesicle-based sample. Background Technology

[0002] Raman spectroscopy has scientific, commercial, and public safety applications. The stability and functionality of lipid nanoparticles (LNPs) depend on their quantity, type, and ratio of individual lipids. The presence of impurities or degraded lipid components significantly affects the quality of LNPs. Current state-of-the-art techniques utilize mass spectrometry and high-performance liquid chromatography (HPLC) methods and systems as analytical tools for LNP analysis. The use of mass spectrometry and HPLC requires significant resources, such as trained users, highly technical instrumentation and data analysis, and substantial time for analyzing the samples used to analyze the LNPs. Summary of the Invention

[0003] According to one aspect of this disclosure, a computer-implemented method on an analytical instrument support device is disclosed. The method includes receiving Raman spectral data associated with compositional training data (e.g., real data), wherein the compositional training data includes data representing one or more components of vesicles. The one or more processors apply one or more preprocessing operations to normalize the Raman spectral data. The one or more processors determine a covariance matrix based at least on the normalized Raman spectral data. The one or more processors determine one or more principal component values ​​associated with the covariance matrix. The one or more processors determine a chemostometric model for one or more parameter levels associated with the data representing one or more components of vesicles based on the one or more principal component values.

[0004] According to another aspect of this disclosure, a computer-implemented method on an analytical instrument support device is disclosed. The method includes obtaining a chemometric model for one or more parameter levels associated with compositional training data, wherein the compositional training data includes data representing one or more components of vesicles. The one or more processors receive Raman spectral data representing at least one Raman spectrum associated with one or more components of a sample of vesicles. The one or more processors transmit the Raman spectral data representing at least one Raman spectrum associated with one or more components of a sample of vesicles to the chemometric model. The one or more processors determine the levels of one or more parameters of one or more components of the sample of vesicles based on the chemometric model.

[0005] According to another aspect of this disclosure, an analytical instrument support system is disclosed. The analytical instrument support system includes one or more computer processors, one or more non-transitory computer-readable storage media, and program instructions stored on at least one or more non-transitory computer-readable storage media, the program instructions being executed by at least one of the one or more processors. The program instructions include: (i) program instructions for receiving Raman spectral data associated with compositional training data, wherein the compositional training data includes data representing one or more components of vesicles; (ii) program instructions for applying one or more preprocessing operations to normalize the Raman spectral data; (iii) program instructions for determining a covariance matrix based at least on the normalized Raman spectral data; (iv) program instructions for determining one or more principal component values ​​associated with the covariance matrix; and (v) program instructions for determining a chemometric model for one or more parameter levels associated with the data representing one or more components of vesicles based on the one or more principal component values.

[0006] According to another aspect of this disclosure, an analytical instrument support system is disclosed. The analytical instrument support system includes one or more computer processors, one or more non-transitory computer-readable storage media, and program instructions stored on at least one or more non-transitory computer-readable storage media, the program instructions being executed by at least one of the one or more processors. The program instructions include: (i) program instructions for obtaining a chemometric model for one or more parameter levels associated with compositional training data, wherein the compositional training data includes data representing one or more components of vesicles; (ii) program instructions for receiving Raman spectral data representing at least one Raman spectrum associated with one or more components of a sample of vesicles; (iii) program instructions for transferring the Raman spectral data representing at least one Raman spectrum associated with one or more components of a sample of vesicles to the chemometric model; and (iv) program instructions for determining the levels of one or more parameters of one or more components of a sample of vesicles based on the chemometric model.

[0007] According to another aspect of this disclosure, an analytical instrument is disclosed. The analytical instrument includes a light source configured to direct light onto a surface of a sample and a spectrometer configured to acquire Raman spectra from the surface of the sample in response to the light source directing light onto the surface of the sample, one or more processors, one or more non-transitory computer-readable storage media, and program instructions stored on at least one of the one or more non-transitory computer-readable storage media, the program instructions being executed by at least one of the one or more processors. Execution of the program instructions by at least one of the one or more processors causes the analytical instrument to perform the following actions: (i) program instructions for receiving Raman spectral data associated with compositional training data, wherein the compositional training data includes data representing one or more components of vesicles; (ii) program instructions for applying one or more preprocessing operations to normalize the Raman spectral data; (iii) program instructions for determining a covariance matrix based at least on the normalized Raman spectral data; (iv) program instructions for determining one or more principal component values ​​associated with the covariance matrix; and (v) program instructions for determining a chemometric model for one or more parameter levels associated with the data representing one or more components of vesicles based on the one or more principal component values.

[0008] According to another aspect of this disclosure, an analytical instrument is disclosed. The analytical instrument includes a light source configured to direct light onto a surface of a sample and a spectrometer configured to acquire Raman spectra from the surface of the sample in response to the light source directing light onto the surface of the sample, one or more processors, one or more non-transitory computer-readable storage media, and program instructions stored on at least one of the one or more non-transitory computer-readable storage media, the program instructions being executed by at least one of the one or more processors. Execution of the program instructions by at least one of the one or more processors causes the analytical instrument to perform the following actions: (i) program instructions for obtaining a chemiostoichiometric model for one or more parameter levels associated with compositional training data, wherein the compositional training data includes data representing one or more components of vesicles; (ii) program instructions for receiving Raman spectral data representing at least one Raman spectrum associated with one or more components of a sample of vesicles; (iii) program instructions for transferring the Raman spectral data representing at least one Raman spectrum associated with one or more components of a sample of vesicles to the chemiostoichiometric model; and (iv) program instructions for determining the levels of one or more parameters of one or more components of a sample of vesicles based on the chemiostoichiometric model.

[0009] According to this disclosure, for some benefits, it is not necessary to require that a system, method, or technique involving measurement-based spectroscopic techniques include all the details characterized herein. Therefore, the specific examples described herein are intended as exemplary applications of the techniques, and alternatives are possible. Attached Figure Description

[0010] The features and advantages of the present technology will become more apparent from the following detailed description of exemplary embodiments in conjunction with the accompanying drawings, wherein: Figure 1 This is a block diagram of an exemplary analysis system according to some embodiments of the present disclosure.

[0011] Figure 2 According to some embodiments of this disclosure, it is shown that Figure 1 The optical architecture of the spectrometer included in the analysis system.

[0012] Figure 3 According to some embodiments of this disclosure, it is shown that Figure 1 Another optical architecture of the spectrometer included in the analysis system.

[0013] Figure 4 According to some embodiments of this disclosure, it is shown that Figure 1 Another optical architecture of the spectrometer included in the analysis system.

[0014] Figure 5 According to some embodiments of this disclosure, it is shown that Figure 1 Another optical architecture of the spectrometer included in the analysis system.

[0015] Figure 6 This is a flowchart illustrating an exemplary process for determining a stoichiometric model based on principal component values ​​of one or more parameter levels associated with data representing vesicle components, according to some embodiments of this disclosure.

[0016] Figure 7 This is a flowchart of an exemplary process for determining the levels of one or more parameters of one or more components of a vesicle sample based on an exemplary stoichiometric model, according to some embodiments of this disclosure.

[0017] Figure 8 This is a diagram illustrating exemplary Raman spectral data of at least one Raman spectrum associated with one or more components of an exemplary vesicle, according to some embodiments of the present disclosure.

[0018] Figure 9 The performance of an exemplary chemostometric model using principal component analysis (PCA) to determine one or more parameters of the quality of one or more components of a sample representing vesicles, according to some embodiments of this disclosure, is demonstrated.

[0019] Figure 10 This is a diagram illustrating exemplary Raman spectral data representing the quantification of single-component lipids in an ethanol solution according to some embodiments of the present disclosure.

[0020] Figure 11 The performance of an exemplary stoichiometric model for determining one or more parameters representing the amount of isopropyl myristate in ethanol, according to some embodiments of the present disclosure, is demonstrated.

[0021] Figure 12 The concentrations of isopropyl myristate-cholesterol were shown according to some embodiments of this disclosure.

[0022] Figure 13 This is a diagram illustrating Raman spectral data representing at least one Raman spectrum associated with one or more components of a vesicle, according to some embodiments of the present disclosure.

[0023] Figure 14 The performance of an exemplary stoichiometric model for determining one or more parameters representing the amount of isopropyl myristate and cholesterol in a mixture with ethanol, according to some embodiments of the present disclosure, is demonstrated.

[0024] Figure 15 The performance of an exemplary stoichiometric model for determining one or more parameters representing the amount of isopropyl myristate and cholesterol in a mixture with ethanol, according to some embodiments of the present disclosure, is demonstrated.

[0025] Figure 16 This is based on some embodiments of the present disclosure. Figure 13 The illustration of the Raman spectral data shows the data excluded from the quantitative model.

[0026] Figure 17 This is a diagram of Raman spectral data according to some embodiments of the present disclosure, which demonstrates the quantification of aqueous pollutants using a partial least squares (PLS) model.

[0027] Figure 18 The performance of an exemplary stoichiometric model for determining one or more parameters representing the quantity of water, according to some embodiments of this disclosure, is demonstrated.

[0028] While this technology is readily adaptable to various modifications and alternatives, specific embodiments have been illustrated by way of example in the accompanying drawings and will be described in detail herein. However, it should be understood that the invention is not limited to the specific forms disclosed. Rather, the purpose of this invention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as defined in the appended claims. Detailed Implementation

[0029] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. In case of any conflict, this document (including the definitions) shall prevail. Exemplary methods and systems are described below; however, similar or equivalent methods and systems may be used in practice or testing with respect to the methods and systems described herein. All publications, patent applications, patents, and other references mentioned herein are incorporated herein by reference in their entirety. The systems, methods, and examples disclosed herein are illustrative only and not limiting.

[0030] As used herein, the terms “comprising,” “including,” “having,” “having,” “may,” “containing,” and their variations are intended to be open-ended transitional phrases, terms, or words that do not exclude the possibility of additional actions or structures. Unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “the” contain plural indicators.

[0031] As used herein, the term “or” is intended to mean inclusive “or” rather than exclusive “or.” That is, unless otherwise specified or clearly apparent from the context, “X uses A or B” means any natural inclusive permutation. That is, if X uses A, X uses B, or X uses both A and B, then “X uses A or B” is satisfied in any of the foregoing examples. Furthermore, unless otherwise specified or clearly apparent from the context that a singular form is involved, the article “a” as used in the subject matter specification and accompanying figures should generally be interpreted as meaning “one or more”.

[0032] The definitions of specific functional groups and chemical terms are described in more detail below. For the purposes of this disclosure, chemical elements are identified according to the CAS version of the periodic table of elements as found on the back cover of the 75th edition of the Handbook of Chemistry and Physics, and the definitions of specific functional groups are generally also based on that document.

[0033] For the numerical ranges listed in this article, each intermediate number with the same precision is explicitly assumed. For example, for the range of 6-9, the numbers 7 and 8 are assumed in addition to 6 and 9; and for the range of 6.0-7.0, the numbers 6.0, 6.1, 6.2, 6.3, 6.4, 6.5, 6.6, 6.7, 6.8, 6.9, and 7.0 are explicitly considered.

[0034] "Raman measurement" refers to a Raman system in which the diameter of the illumination spot remains constant and has a uniform radial distribution.

[0035] "Optical device for generating aspherical diffuse rings" refers to various embodiments for generating distributed light spots, including optical devices for generating aspherical diffuse rings (ADRPO). In some embodiments, the aspherical optics may include so-called axial cone or conical optics that generate intensity rings but have higher-order aspherical terms to generate extended patterns. In some embodiments, the aspherical optics may have coefficients A1 = 0.01, A2 = 0.06, and A4 = 0.002, with all other terms being zero.

[0036] A collimating lens is an optical element that transforms the direction of incident light into a parallel path.

[0037] A "filter" is an optical element that removes certain wavelengths of incident light.

[0038] "Focusing optics" refers to optical elements that transform the direction of incident light into a spatial point.

[0039] "Light source" refers to the light source used for excitation in spectroscopic applications. Exemplary systems and methods may include lasers suitable for Raman spectroscopy, such as 785 nm or 1064 nm. Example light sources may also include broadband sources, such as LEDs. In some embodiments, the laser power affects the values ​​of the baseline and maximum brightness intensity values ​​when scanning a sample. Exemplary systems and methods disclose and envision the use of a wide range of laser power.

[0040] "Sample surface plane" refers to the surface to which the irradiated area of ​​the sample being tested points.

[0041] A "steering mirror" is an optical element used to change the direction of light paths.

[0042] "Raman spectral data" refers to spectra that represent data values ​​for bright and / or dark spectra. Bright spectra are the scattered light from the sample when it illuminates the detector. Dark spectra are the spectra received when no light illuminates the detector. Dark spectra capture the shape of the baseline shift.

[0043] This document discloses systems, methods, and techniques for chemometric modeling applied to determine the levels of one or more parameters of a sample, such as a physical compound or a biological substance. This disclosure is particularly desirable by providing chemometric model determinations that are rapidly available relative to current state-of-the-art methods and systems. For example, this disclosure provides an improved method for determining the levels of one or more parameters of a sample based on at least one Raman spectroscopy of the sample. In some embodiments, this disclosure desirously provides improved methods and systems for determining the levels of one or more parameters of a sample by utilizing continuous Raman spectroscopy scans during lipid formulation, thereby improving accuracy and reducing the analytical time used during current state-of-the-art methods and systems. Improvements of this disclosure include rapid and accurate continuous monitoring during lipid formulation of individual lipid concentrations, degraded lipid impurities, physical properties in solution, and aqueous contamination.

[0044] The improvements of this disclosure can also be used for continuous monitoring of lipid nanoparticle (LNP) formation and for the storage of formulated LNPs for future use. Current state-of-the-art methods and systems require additional time and cost to analyze and / or monitor stored LNPs. The improvements of this disclosure provide a rapid, accurate, and continuous quality control measure by monitoring LNPs to determine the levels of one or more parameters of stored LNPs prior to use.

[0045] As described herein, the Raman measurement parameters of the analytical system are provided to the analytical instrument as initial targets as instructions for obtaining spectra. Raman measurement parameters may include scan time and one or more Raman shift wavenumbers. The systems and methods described herein model these Raman measurement parameters to determine the levels of one or more parameters of one or more components of a vesicle sample based on a chemometric model. For example, a chemometric model is obtained for one or more levels of one or more parameters associated with compositional training data, wherein the compositional training data includes data representing one or more components of the vesicle. This chemometric model determines the levels of one or more parameters of the sample based at least on Raman spectral data of at least one Raman spectrum associated with one or more components of the vesicle sample being fed into the chemometric model.

[0046] Raman spectroscopy is an effective tool for identifying and characterizing various sample compounds and substances. In Raman spectroscopy, light of a known wavelength (usually infrared or near-infrared) from a laser is directly irradiated onto the sample compound or substance. The laser (sometimes called a Raman pump source) interacts with the electron cloud in the molecules of the sample compound or substance, producing a characteristic wavelength shift. The precise nature of this wavelength shift depends on the materials present in the sample compound or substance. Each sample compound or substance produces a unique wavelength characteristic (often called a Raman spectral signature). This unique Raman spectral signature allows for the identification and characterization of the sample compound or substance. More specifically, the spectrum of the light returning from the sample compound or substance is analyzed using a spectrometer to identify the Raman-induced wavelength shift in response to the Raman pump light, and this wavelength signature is then compared to a library of known Raman signatures (e.g., via a computing device), thereby identifying the precise properties of the sample compound or substance.

[0047] This disclosure is now described with reference to the accompanying drawings, wherein like reference numerals are used throughout to refer to like elements. In the following description, numerous specific details are set forth for purposes of explanation in order to provide a more thorough understanding of this disclosure. However, it may be apparent, however, that the systems and methods of this disclosure can be practiced without one or more of these specific details. In other aspects, well-known structures and apparatuses are shown in block diagram form to facilitate the description of the systems and methods of this disclosure.

[0048] It should be understood that although the embodiments described herein are used with a spectrometer or other optical instrument, they can also be constructed as stand-alone devices for measuring the electrochemical properties of sample compounds or substances. Furthermore, while some embodiments are described herein with respect to measuring the electrochemical properties of sample compounds or substances, the exemplary methods and systems described herein can be used to measure other electrochemical properties, such as Raman spectra of sample compounds or substances.

[0049] Exemplary analytical systems can be used in a variety of environments to identify unknown materials, monitor the production and / or formulation of compositions in closed or open environments, assess the threats posed by unknown materials, provide positive identification of packaged raw materials, or provide general safety screening capabilities for various substances. Exemplary analytical systems can include a wide range of sizes, from portable handheld instruments to larger systems permanently installed in laboratories.

[0050] Exemplary Analysis System

[0051] Those skilled in the art will understand that a variety of different optical architectures and arrangements are used in the field of Raman spectroscopy. Figure 1An illustrative example of an analytical system 100 (also referred to herein as “analyzer 100”) is provided, which includes an optical architecture and other components, and is operated to measure one or more Raman spectra from a sample by one or more methods described herein.

[0052] Figure 1 The analyzer 100 shown includes a spectral system 110, which is communicatively coupled to a computing device 120 via a network 130. For example... Figure 1 As shown, the spectral system 110 includes a controller 111, an electronic signal processor 113, and a spectrometer 140 (e.g., a Raman spectrometer).

[0053] It should be understood that in some embodiments, at least a portion of the computing device 120 may be decoupled from the spectral system 110, providing opportunities to increase computing power at a central location or across multiple locations. Those skilled in the art can envision various physical and wireless interconnections between the components of the analysis system 100. It should also be understood that in some embodiments, the spectral system 110 and the computing device 120 may be communicatively coupled without network 130 (e.g., via a dedicated wired or wireless connection). Alternatively, some embodiments of the analyzer 100 may not require the resources of the computing device 120, but may instead utilize the resources within the spectral system 110 to perform the methods described herein. Therefore, the computing device 120 may not be essential for the operation of the analyzer 100 and / or the spectral system 110, and Figure 1 The examples provided should not be considered limiting. As described herein, the analyzer 100 can be used to measure one or more Raman spectra of a sample compound or a plurality of substances using one or more methods described herein.

[0054] It should be understood that in some implementations, Figure 1The analyzer 100 and / or spectroscopic system 110 shown may be housed in a common housing forming an analytical instrument, which may include a benchtop or portable Raman spectrometer device (e.g., a handheld device). However, in other embodiments, one or more components of the analyzer 100 and / or spectroscopic system 110 may be contained in separate housings or devices and may be coupled as needed (e.g., communicative, electrical, mechanical, etc.) to perform the methods described herein. Furthermore, in some embodiments, the operations described herein as being performed by components of the analyzer 100 and / or spectroscopic system 110 may be combined and distributed in various ways. For example, in some embodiments, the electronic signal processor 113 may be part of a controller 111, wherein the controller 111 is configured to perform the operations of the electronic signal processor 113 as described herein. Furthermore, the operations described herein as being performed by the controller 111 may be distributed across multiple controllers. In similar or alternative instances, the operations described herein as being performed by the controller 111 may be distributed among one or more computing devices (e.g., processor 113, computing device 120, or multiple computing devices). In some embodiments, controller 111 is configured to control the operation of spectrometer 140, while processor 113 is configured to control other components of the spectroscopic system 110 (e.g., communication with computing device 120). However, these roles of controller 111 and processor 113 can be combined and distributed in various ways, and in some embodiments, the spectroscopic system 110 includes only controller 111 or processor 113, and the included device performs the functions of both controller 111 and processor 113 as described herein.

[0055] The spectral system 110 may also include additional components (such as a power supply assembly), a user interface 114 (such as a display 112 and / or user input and / or output (“I / O”) devices 109, such as a keyboard, mouse, touchscreen, etc.), optical components (e.g., mirrors, lenses, fiber optic cables, gratings, and filters), etc. The spectrometer 140 included in the spectral system 110 includes one or more optical components 145, a detector 147 (e.g., a CCD detector, a PMT detector, or other detectors known in the art), and a light source 149. The light source 149 directs light onto a sample (not in the optical system). Figure 1 (as shown in the figure) provides an excitation beam (e.g., an excitation laser that provides 785 nm or 1064 nm light).

[0056] As described above, the spectroscopic system 110 and / or spectrometer 140 may comprise a fully integrated portable system operated by a user using battery power to perform Raman spectroscopy measurements in a variety of environments, such as laboratory environments, manufacturing (e.g., bioreactor-based) environments, remote environments, etc. Furthermore, in the same or alternative embodiments, elements of the spectroscopic system 110 may serve as a standalone system operating using a communication connection (e.g., optical, wireless, electrical, mechanical, etc.) and a power outlet connected to a central power source to perform Raman spectroscopy measurements in said various environments.

[0057] Referring now to the light source 149 of the spectrometer 140, it should be understood that the implementation of the light source 149 may emit light of a specific wavelength as required by the application, for example, including or between the following wavelength bands: about 400 nm to about 1064 nm, about 400 nm to about 750 nm, about 400 nm to about 600 nm, about 400 nm to about 500 nm, about 600 nm to about 900 nm, about 700 nm to about 850 nm, about 600 nm to 1064 nm, 750 nm to 1064 nm, 850 nm to 1064 nm, 950 nm to 1064 nm, and a wavelength of about 785 nm, or a wavelength of about 1064 nm.

[0058] Figure 2 provides an illustrative example of one implementation of the optical architecture of optical components including spectrometer 140 (see Figure 1), which are collectively referred to herein as optical system 200. It should be understood that different optical architectures of Raman spectrometers are known in the art, therefore... Figure 2 The examples provided should not be considered limiting. For instance, some implementations employ so-called transmission diffraction instead of reflection diffraction, and there are related differences in optical architecture.

[0059] Figure 2 The illustration shows the light source 149, which serves as the laser component 201 (see Figure 149). Figure 1 One embodiment includes a laser source that generates a beam that propagates along an optical or beam path 230 (e.g., an arrow indicating the direction of beam propagation) to a sample 260. It should be understood that the sample 260 may contain any sample type of interest to the user, including substantially dry samples (e.g., powders, solid materials), substantially fluid samples (e.g., liquids, gases), or combinations of both (e.g., gels). In response to light from the laser assembly 201, the sample 260 generates scattered light (e.g., including Raman and Rayleigh portions of the scattered light) that travels along an optional or beam path 240.

[0060] In some embodiments, the laser component 201 may generate laser power as required by the application, for example, including or within the following power ranges: about 250 mW to about 750 mW; about 250 mW to about 700 mW; about 250 mW to about 650 mW; about 250 mW to about 600 mW; about 250 mW to about 550 mW; about 250 mW to about 500 mW; about 250 mW to about 450 mW; about 250 mW to about 400 mW; about 250 mW to about 350 mW; about 250 mW to about 300 mW; or about 250 mW. Furthermore, in some embodiments, the laser power affects the values ​​of the reference value and the maximum brightness intensity value when scanning the sample 260. It should be understood that other ranges and / or levels of laser power are known in the art, and therefore the examples described with respect to the laser component 201 should not be considered limiting.

[0061] Figure 2 One embodiment of the architecture for directional control of beam paths 230 and 240 and for adjusting one or more characteristics of the beams generated from the laser assembly 201 and the sample 260 is also shown. For example, a steering mirror 202 redirects the beam path 230 to a focusing lens 203, which focuses the beam onto a waveguide phase scrambler 204 (e.g., to adjust the phase characteristics of the beam). The beam exits the waveguide phase scrambler 204 and proceeds to a collimating lens 205 (e.g., which adjusts the collimation characteristics of the beam), and then to a broadband filter 206 that transmits light of a specific wavelength or wavelength range. The beam proceeds to a plane mirror 207, which redirects the beam path 230 to a selective element 209. The selective element 209 may include a dichroic mirror, a notch filter, or other optical elements that have substantially reflective properties for the wavelengths of the beam emitted from the laser assembly 201, while having substantially transmittive properties for the wavelengths or bands of Raman scattered light generated from the sample 260. In the described example, the selective element 209 redirects the beam path 230 to the lens 208, which focuses the beam onto the sample 260. In the described example, the lens 208 may comprise any type of lens known in the art, such as an objective lens for focusing a beam onto the sample 260. Furthermore, as will be described below, some embodiments of the lens 208 include special configurations and features that provide advantages for different types of samples 260.

[0062] In response to the beam from laser assembly 201, lens 208 collects Raman and Rayleigh scattered light generated by sample 260, creating a beam path 240 returning to selective element 209 and second selective element 210. As described above, selective elements 209 and 210 are substantially transmissive to the wavelength of the Raman scattered light, thereby allowing beam path 240 to pass through to additional optical elements that further adjust the path and modulate the characteristics of the beam propagating along beam path 240. For example, the optical elements may include focusing lens 211, plane mirror 212, baffle 213, slit 214, baffle 215, and collimating lens 216.

[0063] The beam path 240 propagates from the collimating lens 216 to the reflecting mirror 220, which reflects the beam path 240 toward the diffraction grating 217. It should be understood that, in Figure 2 In this example, the diffraction grating 217 is a reflective diffraction grating that produces a spectral distribution of light. The beam path 240 then travels to the focusing mirror 219, which redirects the beam path 240 to the focusing lens 221, which guides the beam to the elements of the detector 222. Figure 1 (One implementation of detector 147). It should also be understood that, Figure 2 A baffle 218 for controlling stray light is shown in some embodiments.

[0064] As described above, it should be understood that various different lens 208 embodiments can be obtained, providing different focusing and light-collecting characteristics. For example, Figure 3 An example implementation of an optical architecture for analyzing samples contained in packaging (e.g., bags, bottles, etc.) is provided, wherein the optical architecture includes some components of an optical system 200 (see [link to implementation details]). Figure 2 ) and has lens 208 (see Figure 2 Other components that define the optical device 300 are collectively referred to as optical device 300. In the described example, optical device 300 includes element 302, which may include focusing lens 203 (see [link to description]). Figure 2 The output may be from a light source 149 or a laser assembly 201 or a Raman laser 119, 519—see [link to relevant documentation]. Figure 1 , Figure 2 and Figure 5The light beam is guided to collimating lens 304, which produces a substantially collimated beam. In the example described, collimating lens 304 can be movably mounted such that its position can be changed along the axis of the optical path. The range of motion includes approximately 0.1 mm to approximately 10 mm to allow the spot size on the sample surface to vary from approximately 10 μm to approximately 10 mm. It should also be understood that in some embodiments, any one (alone or in combination) of collimating lens 304, concave focusing lens 312, and / or focusing optics 314 can be movably mounted to achieve variations in spot size.

[0065] Collimating lens 304 guides the substantially collimated beam to optics 308, which generates an aspherical diffuse ring, to produce a radially diffused light pattern. The intensity of the output from optics 308 is stronger at the outer edges of the resulting pattern than at the center. While this pattern can be directly projected onto the sample surface 316, in practical applications, it is advantageous to use one or more steering mirrors 310, one or more filters 306, and focusing elements such as a concave focusing lens 312 and a focusing optics 314 to guide the radially diffused light pattern onto the sample surface 316.

[0066] Figure 4 Lens 208 is provided (see Figure 2 Another example of an implementation of the method, which can be used to analyze fluid or semi-fluid samples. Figure 4 The illustrated implementation includes some components of the optical system 200 and other components that provide features commonly referred to as “immersion probes”, wherein these components are collectively referred to herein as optical device 400. Figure 4 The illustrated embodiment includes a spherical lens 440 disposed within a cylindrical probe tip 410 at a lens opening 418. A seal between the probe tip 410 and the lens 440 is formed at the opening by any means known in the art, including all forms of welding or sintering and the use of epoxides or other adhesives. The probe tip 410 may have any length. Optionally, the probe tip 410 may have threads 414 on its inner surface and may be extended using a probe tube 430 having a threaded collar 432 for screwing into the probe tip 410. Optionally, a seal is formed between the probe tube lip 437 and the distal end of the probe tip 410. Furthermore, in this example, the optical device 400 includes an optical fiber coupler 439 transmitting from the laser assembly 201 (see [link to example]). Figure 2 The light source irradiated by the lens 440 and the scattered light from the sample 260 (see Figure 2), wherein the sample 260 may include a liquid sample, and the lens 440 is immersed in the liquid. Additionally, in this example, the optical device 400 may be configured to work with the spectral system 110 (see Figure 2). Figure 1The separated components enable optical communication between the spectral system 110 and the optical device 400 via optical fiber.

[0067] It should be understood that Figure 3 and Figure 4 The examples provided are for illustrative purposes only; in practical applications, some implementations may include or omit components as needed. For instance, in some implementations, in applications employing fiber optic coupling or requiring beam adjustment / internal environment protection, one or more windows, collimating lenses, or other optical elements may be added. Therefore, Figure 3 and Figure 4 The examples provided should not be considered restrictive.

[0068] Figure 5 Provided including spectrometer 140 (see Figure 1 This is another example of an implementation of the optical architecture of the optical components, collectively referred to herein as optical system 500. It should be understood that different optical architectures of Raman spectrometers are known in the art, and therefore, with... Figures 1 to 4 Similar examples exist. Figure 5 Examples should not be considered restrictive.

[0069] Figure 5 An example is shown as a light source 149 for Raman laser 119 (see Figure 1 One embodiment includes a laser source that generates a beam that propagates along a first optical or beam path 510 (e.g., an arrow indicating the direction of beam propagation) to sample 530. It should be understood that, with reference to sample 260 (see...),... Figure 2 Similarly, sample 530 may contain any sample type of interest to the user, including substantially dry samples (such as powders, solid materials), substantially fluid samples (such as liquids, gases), or combinations of both (such as gels). In response to light from Raman laser 119, sample 530 generates scattered light (e.g., including Raman and Rayleigh portions of the scattered light) that travels along a second optical or beam path 520.

[0070] In some embodiments, the Raman laser 119 may generate laser power as required by the application, for example including or ranging from about 250 mW to about 1050 mW, including various sub-ranges therein, such as the non-limiting sub-ranges described above with respect to the light source 149 and the laser assembly 201. It should also be understood that in some embodiments, the laser power affects the values ​​of the reference value and the maximum brightness intensity value when scanning the sample 530.

[0071] Figure 5An architecture for directional control of a first beam path 510 and / or a second beam path 520 is illustrated in some embodiments. In some embodiments, beam paths 510, 520 may be controlled by one or more optical elements, such as a steering mirror, a waveguide phase scrambler, various lenses, broadband filters, or selective elements (e.g., mirrors, notch filters, or other elements that have fundamental reflective properties for the wavelength of the beam emitted by the Raman laser 119 and / or fundamental transmission properties for the wavelength or band of the Raman scattered light generated by the sample 530). In the described example, selective element 511 may transmit the laser wavelength emitted from the Raman laser 119, thereby allowing the first beam path 510 to be directed to a lens 508 that focuses the beam onto the sample 530. In the described example, lens 508 may comprise any lens type known in the art, such as an objective lens or such as... Figure 3 and Figure 4 The lens architecture used in the optional devices 300 or 400 shown can focus the light beam onto the sample 530.

[0072] Some embodiments of lens 508 have special configurations and characteristics that can provide advantages for different types of samples. For example, lens 508 can collect Raman and Rayleigh scattered light generated by sample 530 after being excited by the Raman laser beam 119. The scattered light collected by lens 508 is reflected back from the surface of sample 530 and returns along a first beam path 510 to a selective element 511 (e.g., a beam splitter, such as a dichroic mirror), which guides the scattered light along a second beam path 520. In some embodiments, selective element 511 has a fundamentally reflective characteristic to the wavelength of the Raman scattered light, allowing the second beam path 520 to be guided to other optical elements to further adjust the path and modulate the characteristics of the beam propagating along that path. Other optical devices may also be considered for selective element 511 to guide the scattered light along the second beam path 520.

[0073] like Figure 5 As shown, the optical system 500 also includes one or more optical components 115 (also referred to herein as optical components 115a-115c), which may include elements such as collimating lenses and mirrors, filters (e.g., notch filters), diffraction gratings, and / or mirror relay systems. Scattered light is guided by one or more optical components 115a-115c to detector 117. Figure 1(One embodiment of detector 147). Signal processing and / or digitization of the signal associated with the scattered light received by detector 117 is performed by an electrical signal processor associated with optical system 500, which may be, for example, processor 113, controller 111, computing device 120, or a combination thereof. For example, in some embodiments, electrical signal processor 113 may be a suitably programmed microprocessor or application-specific integrated circuit that includes read-only or read-write memory of any known type for storing instructions and data required for the operation of the spectrometer described herein.

[0074] As described above, it should be understood that various different lens 508 embodiments are available that provide different focusing and light-collecting characteristics.

[0075] Return to Figure 1 The computing device 120 can be a standalone device, server, Internet of Things (IoT) device, laptop computer, tablet computer, netbook computer, personal computer (PC), smartphone, personal digital assistant (PDA), desktop computer, or any programmable electronic device capable of receiving, sending, and processing data. In some embodiments, the computing device 120 includes one or more processors, one or more input / output processors, and one or more memory or data storage devices. In some implementations, the computing device 120 also includes one or more input / output devices, such as a display, touchscreen, keyboard, mouse, etc., which can be used to provide calibration or setting options for the user to operate the spectral system, provide the user with analytical results, or a combination thereof.

[0076] Similarly, controller 111 (see Figure 1 The controller 111 may include an electronic processor, input / output (I / O) interfaces, and data storage devices (not shown); however, it should be understood that the controller 111 may have additional or fewer components. The controller 111 is adapted to the stated application scenarios and setup conditions, such as being configurable with multiple electronic processors, multiple input / output interfaces, multiple data storage devices, or combinations thereof. In some embodiments, some or all of the components included in the controller 111 may be attached to one or more motherboards and enclosed in a housing (e.g., including plastic, metal, and / or other materials). In some embodiments, some of these components may be fabricated onto a single system-on-a-chip (SoC) (e.g., an SoC may include one or more processing devices and one or more storage devices).

[0077] As used herein, "processor" or "electronic processor" or "electronic signal processor" means any device or part of a device that processes electronic data from registers and / or memory to convert electronic data that can be stored in registers and / or memory. Processing device 111 may include one or more digital signal processors (DSPs), application-specific integrated circuits (ASICs), central processing units (CPUs), graphics processing units (GPUs), cryptographic processors (dedicated processors that execute cryptographic algorithms within hardware), server processors, or any other suitable processing device.

[0078] The data storage device included in controller 111 may include one or more local or remote memory devices, such as random access memory (RAM) (e.g., static RAM (SRAM), magnetic RAM (MRAM), dynamic RAM (DRAM), resistive RAM (RRAM), or conductive bridged RAM (CBRAM), hard disk drive-based memory, solid-state memory, network drive, cloud drive, or any combination of memory devices. In some embodiments, the data storage device may include memory sharing a die with the processor. In such embodiments, the memory may be used as cache memory and may include, for example, embedded dynamic random access memory (eDRAM) or spin-transfer torque magnetic random access memory (STT-MRAM). In some embodiments, the data storage device may include a non-transitory computer-readable medium having instructions thereon that, when executed by one or more processors (e.g., an electronic processor included in controller 111), cause controller 111 to store various applications and data for performing one or more of the methods described herein or portions thereof. For example, one or more data storage devices may store modeling programs, parameter and component data, chemometric model data, composition training data, or combinations thereof. Alternatively or additionally, the program and associated data, or portions thereof, may be stored on and executed on the computing device 120. Furthermore, it should be understood that each method described herein can be implemented via one or more applications.

[0079] In some implementations, the modeling procedure can be configured to execute one or more sets of program instructions to perform one or more operations and / or processes, such as accessing, retrieving, receiving, identifying, analyzing, determining and / or generating parametric component data, chemometric model data and / or composition training data.

[0080] In some embodiments, the parameter component data may include data on one or more exemplary components associated with the vesicle. In some embodiments, the vesicle may be lipid nanoparticles (LNPs). Lipid nanoparticles (LNPs) may also include one or more exemplary components. The one or more exemplary components may include one or more solvents, one or more lipids, one or more surfactants, one or more LNPs, one or more micelles, one or more stabilizers (e.g., humectants such as glycerol or propylene glycol), one or more buffers, one or more salts, one or more waxes, and / or one or more intermediate phases of the LNP stored on a data storage device.

[0081] In some embodiments, parameter and component data may include data representing one or more parameters of the quality of one or more components of vesicles stored on a data storage device. In some embodiments, one or more parameters representing the quality of one or more exemplary components may indicate the identity of one or more exemplary components modeled by a stoichiometric model, as described herein. In some embodiments, the identity associated with one or more exemplary components may include one or more solvents, one or more lipids, one or more surfactants, one or more LNPs, one or more micelles, one or more stabilizers (e.g., humectants such as glycerol or propylene glycol), one or more buffers, one or more salts, one or more waxes and / or one or more intermediate phases of LNPs, or combinations thereof, as will be described in further detail below.

[0082] In some embodiments, the parameter component data may include data of one or more parameters representing the quantity of one or more exemplary components of vesicles stored on a data storage device. In some embodiments, the one or more parameters representing the quantity of one or more exemplary components of vesicles may indicate the amount of one or more exemplary components modeled by a stoichiometric model, as described herein. In some embodiments, the amount of one or more exemplary components may include concentrations such as mass concentration, molar concentration, volume concentration, and number concentration. In some embodiments, the amount of one or more exemplary components may include a ratio of one or more components of vesicles.

[0083] In some embodiments, the quantities associated with one or more parameters may indicate that one or more components of the vesicle have been degraded and / or that the vesicle has been contaminated (e.g., aqueous contamination). In some embodiments, the quantities associated with one or more parameters may indicate the concentration of each individual lipid, each individual solvent, each individual surfactant, each individual LNP, each individual micelle, or each individual intermediate phase of the LNP. In some embodiments, the quantities associated with one or more parameters may indicate one or more degraded lipid impurities. In some embodiments, the quantities associated with one or more parameters may include the physical properties (e.g., monomers and aggregates) of any one or any combination of solvents, lipids, surfactants, LNPs, micelles, intermediate phases of LNPs, or any combination thereof. In some embodiments, the quantities associated with one or more parameters may indicate aqueous contamination of the vesicle.

[0084] In some implementations, chemometric model data may include any number of regression analysis or machine learning procedures stored on a data storage device, including partial least squares regression (PLS) models, partial least squares discriminant analysis (PLSDA), principal component analysis (PCA) models, principal component regression (PCR) models, least absolute shrinkage and selection operator (LASSO) models, elastic network regression models, support vector machine (SVM) models, neural network models, or combinations thereof. A brief description of the use of these regression analysis or machine learning procedures is described below.

[0085] In some implementations, the regression analysis and / or machine learning processes described above can be implemented on one or more processors. These processors may be included in the controller 111 and / or in a third-party computing device (e.g., computing device 120) communicatively connected to the spectral system 110.

[0086] In some implementations, PCA is an unsupervised statistical model, also known as singular value decomposition. It can learn to model the training dataset on the test dataset, performing outlier detection on these principal components to find which samples belong to the same distribution as the training dataset. Therefore, this can be a uniclass classification model. Principal components can be computed by performing eigenvalue decomposition on the covariance matrix of the data. The eigenvectors with the highest corresponding eigenvalues ​​represent most of the variance in the data. This creates an orthogonal space in which the data can be represented. The main hyperparameter here is the number of eigenvectors used to represent the data. kUsing more eigenvectors will give the model a higher explained variance. Outliers can then be identified using two statistical tests: the Hotelling T2 test and the Q-residual test. The Hotelling T2 test focuses on the distance of a sample to the remaining samples in the principal component space, while the Q-test focuses on the residual between the sample and the reconstructed sample after being transformed to and back into PC space. These tests are complementary, and if either test classifies a sample as an outlier, in some implementations, the system disclosed herein can treat that sample as an outlier. Because PCA is a dimensionality reduction algorithm, it can also be used as a preprocessing step for other models. The reduced dimensionality may lead to a decrease in overfitting to the training data.

[0087] In some implementations, PLS, or Partial Least Squares Regression (also known as "Latent Structure Projection"), is a statistical method that summarizes and combines features from principal component analysis and multiple regression. It can be used to predict a set of dependent variables from a very large set of independent variables (i.e., predictor variables). The goal of PLS ​​regression is to predict Y and X and describe their common structure. When Y is a vector and X is full rank, this goal can be achieved using ordinary multiple regression. When the number of predictor variables is large compared to the number of observations, X may be singular, and the regression method is no longer feasible (i.e., due to multicollinearity).

[0088] In some implementations, PLSDA is an adaptation of PLS ​​for categorical target variables. The procedure here is similar to PCA, in terms of dimensionality reduction to obtain scores and loadings; however, for PLS, the decomposition is performed in a way that maximizes the covariance between the predictor variable and the target variable across these scores. On the scores, a regression algorithm can be trained to predict the predictor variable. In PLSDA, the target variable is given as a one-hot encoded vector, against which regression can be computed.

[0089] In some implementations, SVM is used for binary classification, where a choice is made between two classes. SVM is linear and attempts to construct a hyperplane in the feature space that maximizes the separation of training data points based on their class. Classification then involves checking which side of the hyperplane a new test point lies on and assigning it the corresponding class. SVM can increase computational power by using kernel functions. These kernel functions allow for non-linear transformations, meaning a non-linear decision surface can be constructed. Each kernel function has its own set of hyperparameters, allowing for further tuning of the model. While the basic SVM is used for binary classification, it can be extended to also allow multi-class classification. In some implementations, PCA decomposition can be performed before the SVM to prevent or limit overfitting, as described above. SVM can also be used as a single-class model for outlier detection. In this case, the SVM is trained on a dataset containing only samples of the class to be identified. A minimum envelope is then constructed around this dataset in the feature space as a hyperplane. Any new test point outside the envelope is classified as an outlier. This model can be used as a standalone single-class model for authentication, or as an outlier model in addition to a multi-class classifier. In some implementations, dimensionality reduction may not be used for single-class SVMs. Such single-class SVMs perform well on high-dimensional data in the system disclosed herein without using PCA for feature extraction.

[0090] In some implementations, LASSO, or the minimum absolute shrinkage and selection operator, is a statistical formula used for regularization and feature selection in data models. It is superior to regression methods used for more accurate predictions. The model uses shrinkage, where data values ​​are shrunk towards the center point, which serves as the mean. The LASSO procedure provides simple, sparse models (i.e., models with fewer parameters). This particular type of regression is well-suited for models exhibiting high levels of multicollinearity or for automating certain types of model selection, such as variable selection / parameter elimination.

[0091] In some implementations, the resilient network approach overcomes the limitations of the LASSO method, which uses a penalty function based on Equation 1, as follows: (1) As shown above, the use of Equation 1 has several limitations. In some implementations, when provided with high-dimensional data with a small number of instances, LASSO selects at most n variables before saturation. Furthermore, if a set of highly correlated variables exists, LASSO tends to select one variable from the set and ignore the others. To overcome these limitations, elastic networks add a quadratic component to the penalty, which, when used alone, is ridge regression (also known as Tikhonov regularization).

[0092] In some embodiments, compositional training data may include any number of Raman spectra, parameters, components, vesicles, one or more experimental designs, or combinations thereof, stored on a data storage device. In some embodiments, multiple compositional training data are used to train any one of one or more chemosimetric models, as described herein. In some embodiments, compositional training data is stored on compositional training data.

[0093] In some embodiments, one or more experimental designs may include uniform designs, factorial designs, partial factorial designs, response surface methodology, randomized designs, block designs, Box-Wilson central composite designs, Box-Behnken designs, or combinations thereof. In some embodiments, the experimental design may include a uniform design.

[0094] In some implementations, one or more experimental designs are provided as a process for using training data to conduct research in an objective and controlled environment. Precision is maximized, and specific conclusions can be drawn for one or more hypothesis statements. One or more experimental designs can determine the effect of one or more factors or independent variables on one or more dependent variables.

[0095] In some implementations, the I / O interface of controller 111 may include one or more communication chips, connectors, and / or other hardware and software to manage communication between controller 111 and other components. The I / O interface may include interface circuitry for coupling to one or more components using any suitable interface, such as a Universal Serial Bus (USB) interface, a High Definition Multimedia Interface (HDMI) interface, a Controller Area Network (CAN) interface, a Serial Peripheral Interface (SPI), an Ethernet interface, a wireless interface, or any other suitable interface. For example, the I / O interface may include circuitry for managing wireless communication for transmitting data to and from controller 111. The term "wireless" and its derivatives can be used to describe circuits, devices, systems, methods, techniques, communication channels, etc., that transmit data over non-solid media using modulated electromagnetic radiation. This term does not imply that the associated device contains no wires, although in some specific implementations the associated device may contain no wires. The circuitry included in the I / O interface for managing wireless communications may implement any of a number of wireless standards or protocols, including but not limited to Institute of Electrical Engineers (IEEE) standards, including Wi-Fi (IEEE 802.11 series), IEEE 802.16 standards (e.g., IEEE 802.16-2005 amendments), Long Term Evolution (LTE) projects, and any amendments, updates, and / or revisions (e.g., Advanced LTE projects, Ultra Mobile Broadband (UMB) projects (also known as “3GPP2”), etc.). In some implementations, the circuitry included in the I / O interface for managing wireless communications may operate according to Global System for Mobile Telecommunications (GSM), General Packet Radio Service (GPRS), Universal Mobile Telecommunications System (UMTS), High Speed ​​Packet Access (HSPA), Evolved HSPA (E-HSPA), or LTE networks. In some implementations, the circuitry included in the I / O interface for managing wireless communications may operate according to enhanced data for GSM Evolution (EDGE), GSM EDGE Radio Access Network (GERAN), Universal Terrestrial Radio Access Network (UTRAN), or Evolved UTRAN (E-UTRAN). In some implementations, the circuitry included in the I / O interface for managing wireless communications may operate according to Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Digital Enhanced Cordless Telecommunications (DECT), Evolved Data Optimization (EV-DO) and its derivatives, as well as any other wireless protocol designated as 3G, 4G, 5G, or higher. In some implementations, the I / O interface may include one or more antennas (e.g., one or more antenna arrays) for receiving and / or transmitting wireless communications.

[0096] In some embodiments, the analysis system 100 provides a standalone or dedicated analytical instrument or device (or group of instruments or devices) configured to perform short scans and analyses of samples. However, in other embodiments, the analysis system 100 may be configured to perform additional scans or analyses of one or more samples. Combining these scanning or analytical capabilities within a single system (e.g., a single analytical instrument) achieves the desired efficiency and improves analytical accuracy because samples can be scanned multiple times without changing sample location, reconfiguring the analytical instrument, or performing separate scans of other samples combined with the target sample, all of which can lead to delays and potential contamination or unexpected discrepancies between scans.

[0097] Exemplary materials

[0098] The exemplary systems and methods described herein may include a variety of materials, such as vesicles and various components of exemplary vesicles. Exemplary vesicles include one or more exemplary components, such as liposomes, solid lipid nanoparticles (LNPs), or non-layered lipid nanoparticles (e.g., inverted cubic and hexagonal phases), each of which is discussed in more detail below.

[0099] A. Exemplary Components

[0100] Exemplary components may include solvents, lipids, surfactants, lipid nanoparticles (LNPs), micelles, stabilizers (e.g., humectants such as glycerol or propylene glycol), buffers, salts, waxes, and / or intermediate phases, each of which is discussed below.

[0101] Exemplary lipid nanoparticles (LNPs) or liposomes may include isopropyl myristate and / or cholesterol.

[0102] 1. Exemplary solvents

[0103] Exemplary solvents may include alcohols, ethers, sulfoxides, hydrocarbons, aromatic compounds, and halogenated compounds (e.g., CHCl3 and CCl4). In some embodiments, exemplary solvents may include ethanol.

[0104] The exemplary solvent does not include water, where water represents aqueous contamination of the formulated vesicles.

[0105] 2. Exemplary lipids

[0106] Exemplary lipids can improve nanoparticle properties, including particle stability, delivery efficacy, tolerability, biodistribution, or combinations thereof.

[0107] In some embodiments, exemplary lipids may include triglycerides, fats, phospholipids, cationic lipids, PEGylated lipids, bioionizable lipids, phospholipids, cholesterol, waxes, steroids, or any combination thereof.

[0108] In some implementations, exemplary lipids may include cholesterol, isopropyl myristate, and stearate.

[0109] In some embodiments, exemplary lipids may include phosphatidylcholine and phosphatidylethanolamine.

[0110] In some embodiments, exemplary lipids may include (3060ilo), tetra(8-methylnonyl)-3,3',3'',3'''-(((methylazonyl)-bis(propane-3,1-diyl))bis(azonyltriyl))tetrapropionate; (9A1P9), (2-(dioctylammonium)ethyl)decyl phosphate; (A2-Iso5-2DC18), ethyl 5,5-di((Z)-heptadec-8-en-1-yl)-1-(3-(pyrrolidone-1-yl)propyl)-2,5-dihydro-1H-imidazolium-2-carboxylate; (ALC-0315), ((4-hydroxybutyl)azadiyl)bis(hexane-6,1-diyl)bis(2-hexyldecyl) Ester); (ALC-0159), 2-[(polyethylene glycol)-2000]-N,N-tetracosylacetamide; (β-sitosterol), (3S,8S,9S,10R,13R,14S,17R)-17-((2R,5R)-5-ethyl-6-methylheptane-2-yl)-10,13-dimethyl-2,3,4,7,8,9,10,11,12,13,14,15,16,17-tetradecano-1H-cyclopenta[a]phenanthrene-3-ol; (BAME-O16B), 3,3'-((3-methyl-9-oxo-10-oxa-13,14-dithia-3,6-diazahexadecyl)azadiyl)di bis(2-(dodecyl dithioalkyl)ethyl) propionate; (BHEM-cholesterol), 2-(((((3S,8S,9S,10R,13R,14S,17R)-10,13-dimethyl-17-((R)-6-methylheptane-2-yl)-2,3,4,7,8,9,10,11,12,13,14,15,16,17-tetradecanohydro-1Hcyclopentan[a]phenanthrene-3-yl)oxy)carbonyl)amino)-N,N-bis(2-hydroxyethyl)-N-methylethane-1-ammonium bromide; (C12-200), 1,1'-((2-(4-(2-((2-(bis(2-hydroxydodecyl)amino)ethyl)(2-hydroxy (dodecyl)amino)ethyl)piperazine-1-yl)ethyl)azadiyl)bis(dodecyl-2-ol); (cKK-E12), 3,6-bis(4-(bis(2-hydroxydodecyl)amino)butyl)piperazine-2,5-dione; DC-cholesterol, 3β-[N-(N′,N′-dimethylaminoethane)-carbamoyl]cholesterol; (DLin-MC3-DMA), (6Z,9Z,28Z,31Z)-heptadec-6,9,28,31-tetraen-19-yl 4-(dimethylamino)butyrate; (DOPE), 1,2-dioleoyl-sn-glycerol-3-phosphate ethanolamine; (DOSPA), 2,3-dioleoyloxy- N -[2-(Speramido)ethyl]- N , N-Dimethyl-1-propanetrifluoroacetate; (DOTAP), 1,2-dioleoyl-3-trimethylammonium-propane; (DOTMA), 1,2-di-O-octadecenyl-3-trimethylammonium-propane; (DSPC), 1,2-distearyl-sn-glycerol-3-phosphocholine; ePC, ethylphosphatidylcholine; (FTT5), hexa(octyl-3-yl)9,9',9'',9''',9''',9''''-((((benzene-1,3,5-tricarbonyl)tri(azadiyl))tri(propane-3,1-diyl))tri(azatriyl))hexanoate; (lipid H (SM-102)), heptadecan-9-yl 8-((2-hydroxyethyl)(6-oxo-6-(undecyloxy)hexyl)amino)octanoate; (OF-Deg-Lin), (((3,6-dioxopiperazine-2,5-diyl)bis(butane-4,1-diyl))bis(azanetriyl))tetra(ethane-2,1-diyl) (9Z,9′Z,9″Z, Z,12Z,12′Z,12″Z, Z)-tetra(octadecyl-9,12-dienoate); (PEG2000-DMG), 1,2-dimyristoyl rac-glycerol-3-methoxy polyethylene glycol-2000; TT3,N1,N3,N5-tris(3-(docodecylamino)propyl)benzene-1,3,5-tricarboxamide; or combinations thereof.

[0111] In some respects, exemplary lipids may include (DSPC), 1,2-distearate-sn-glycerol-3-phosphocholine and / or ePC, ethylphosphatidylcholine.

[0112] 3. Exemplary intermediate phase

[0113] Exemplary intermediate phases of LNPs may include layered phases, hexagonal phases, bicontinuous cubic phases, lipid nanoparticles (LNPs), or combinations thereof.

[0114] Exemplary operating method

[0115] Now for reference Figure 6 The flowchart illustrates a process 600 for determining a stoichiometric model according to some embodiments of this disclosure. As described above, process 600 can be implemented using a spectroscopic system 110. Process 600 is described herein as being executed via a controller 111. However, it should be understood that process 600 can be executed by one or more software and / or hardware components in various combinations and configurations. Figure 6 As shown, process 600 may include operations 602, 604, 606, 608, and 610. In some embodiments, process 600 is performed as follows: Figure 6 The order shown in the image is as follows.

[0116] In operation 602, controller 111 receives Raman spectral data associated with compositional training data, which includes data representing one or more components of a vesicle.

[0117] In operation 604, controller 111 applies one or more preprocessing operations to normalize the Raman spectral data.

[0118] In some embodiments, controller 111 normalizes Raman spectral data representing one or more components of vesicles by applying one or more preprocessing operations to the Raman spectral data. In some embodiments, the one or more preprocessing operations include region selection, spectral averaging, convolution filtering, first derivative, second derivative, standard normalization, multiplicative scattering correction, background removal, or any combination thereof. One or more normalized Raman spectra associated with the constituent training data are used to determine a chemostometric model, as described below.

[0119] In some embodiments, controller 111 applies region selection preprocessing to Raman spectral data representing one or more components of vesicles. In some embodiments, a portion of the entire Raman spectral data is used during region selection. Using only a portion of the entire Raman spectral data may have advantages in certain respects. For example, in some respects, very high and very low wavenumber regions of Raman spectral data typically have very low signal-to-noise ratios (SNR), which may present limited relevance, and training on noisy data may lead to overfitting. In another instance, in some respects, different substances can be distinguished based on different regions of Raman spectral data where specific peaks (i.e., wavenumbers) can be observed. In such respects, the remainder of the Raman spectral data may be observed to be less relevant, and data representing the remainder of the Raman spectral data may be discarded. In some embodiments, region selection is a hyperparameter.

[0120] In some implementations, controller 111 applies at least a second preprocessing operation, which includes a standard normal variable (SNV). During SNV scaling, each spectral data point is scaled using a standard normal transformation, as shown in equation (2) below: (2) Where, x i It is the first in the Raman spectral data i There are _ data points, where µ is the average intensity of the spectrum, σ is the standard deviation of the intensity, and x_i is the average intensity of the spectrum. iSNV It is x i The correction value.

[0121] In operation 606, controller 111 determines the covariance matrix based at least on normalized Raman spectral data.

[0122] The controller 111 determines the covariance matrix by generating a P×P symmetric matrix using normalized Raman spectra. The covariance matrix can be a P-dimensional dataset, such as a 2-dimensional dataset, or a 2048×2048 symmetric matrix with 2048 variables.

[0123] In operation 608, controller 111 determines one or more principal component values ​​associated with the covariance matrix.

[0124] In some implementations, controller 111 determines one or more eigenvectors and one or more eigenvalues ​​based on the covariance matrix. The eigenvectors and eigenvalues ​​are used to determine one or more principal component values. Controller 111 determines one or more eigenvectors associated with the covariance matrix, at least based on the orientation of the axis at the point of maximum variance in the Raman spectral data. The eigenvectors are commonly referred to as principal component values. Controller 111 determines one or more eigenvalues, where the eigenvalues ​​are coefficients associated with the eigenvectors, and the one or more eigenvalues ​​can provide the amount of variance carried in each principal component value. Controller 111 organizes the eigenvectors from highest to lowest based at least on one or more eigenvalues, resulting in the principal component values ​​being sorted according to their importance. In some implementations, controller 111 determines the percentage of variance by dividing each eigenvalue by the sum of the eigenvalues.

[0125] In operation 610, controller 111 determines a stoichiometric model based on one or more principal component values ​​for one or more parameter levels associated with data representing one or more components of vesicles.

[0126] In some embodiments, controller 111 determines a chemometric model, which includes a partial least squares regression (PLS) model, principal component analysis (PCA), principal component regression (PCR) model, least absolute shrinkage and selection operator (LASSO) model, elastic network regression model, support vector machine (SVM) model, or neural network model. In some embodiments, the chemometric model includes a PLS model and / or a PCA model.

[0127] In some implementations, controller 111 determines a chemostometric model based on one or more principal component values. Controller 111 determines the eigenvectors of the chemostometric model by generating a matrix with columns of eigenvectors as defined above. Based at least on the eigenvectors, Raman spectral data representing one or more components of the vesicles are redirected from their original axes to axes represented by one or more principal component values. Therefore, a chemostometric model is determined for one or more parameter levels associated with data representing one or more components that constitute the training data.

[0128] In some aspects of process 600, one or more parameter levels represent the quality of one or more components of a vesicle, wherein one of the levels indicates the identity of one or more components of a vesicle modeled by a stoichiometric model as described above.

[0129] In some aspects of process 600, one or more parameter levels represent the amount of one or more components of a vesicle, wherein one of the levels indicates the amount of one or more components of a vesicle modeled by a stoichiometric model as described above.

[0130] In some aspects, operation 602 may also include: controller 111 analyzing the compositional training data and identifying (i) one or more components, (ii) one or more preprocessing operations associated with each corresponding compositional training data, (iii) one or more parameters representing the quality of one or more components of a vesicle sample, and (iv) one or more parameters representing the quality of one or more components of a vesicle.

[0131] In some aspects, operation 602 may also include compositional training data having one or more parameter levels representing the quality of one or more components of a vesicle. The compositional training data includes one or more parameter levels representing the quantity of one or more components of a vesicle. Multiple compositional training data include any one or more components, such as solvents, lipids, surfactants, lipid nanoparticles (LNPs), micelles, intermediate phases of LNPs, or any combination thereof. In some embodiments, the lipid nanoparticles or liposomes include isopropyl myristate and cholesterol.

[0132] The constituent training data includes any one or more experimental designs. Experimental designs include uniform designs, factorial designs, partial factorial designs, response surface methodology, randomized designs, block designs, Box-Wilson center composite designs, Box-Behnken designs, or combinations thereof. In some implementations, multiple constituent training data include uniform designs. One or more experimental designs are provided as a process for constituent training data used in a study conducted in an objective and controlled environment. Precision is maximized, and specific conclusions can be drawn for one or more hypothesis statements. One or more experimental designs can determine the effect of one or more factors or independent variables on one or more dependent variables. The constituent training data is generated based on any one of the one or more experimental designs, wherein the constituent training data is generated during a process in an objective and controlled environment.

[0133] Now for reference Figure 7The flowchart illustrates a process 700 for determining the levels of one or more parameters of a sample of vesicles according to some embodiments of this disclosure. As described above, process 700 can be implemented using a spectral system 110. Process 700 is described herein as being executed via a controller 111. However, it should be understood that process 700 can be executed by one or more software and / or hardware components in various combinations and configurations. Figure 7 As shown in FIG. 7, process 700 may include operations 702, 704, 706, and 708. In some embodiments, process 700 is executed in the order shown in FIG. 7.

[0134] In operation 702, controller 111 obtains a stoichiometric model of one or more parameter levels associated with composition training data, wherein the composition training data includes data representing one or more components of vesicles.

[0135] In some embodiments, process 700 may further include obtaining a stoichiometric model by controller 111, wherein controller 111 receives compositional training data representing one or more components of vesicles. Controller 111 determines a stoichiometric model of one or more parameter levels associated with the compositional training data.

[0136] In operation 704, controller 111 receives Raman spectral data representing at least one Raman spectrum associated with one or more components of a sample of a vesicle.

[0137] In some implementations, operation 704 can be initiated by scanning one or more samples of the vesicles, as discussed in detail above. The sample is scanned using at least the spectral system 110, as described above. Figure 1-5 As described above, the spectral system 110 directs a Raman laser beam (e.g., light) as described above onto the surface or focal point of the vesicle sample. For example, the Raman laser beam can be directed toward a container containing lipid vesicles, wherein the lipid vesicles comprise one or more components, including solvents, lipids, surfactants, lipid nanoparticles (LNPs), micelles, or intermediate phases of lipid nanoparticles (LNPs). The resulting scattered light is directed back through the selection element 511, and the scattered light travels along the second beam path 520 and through the optical assembly 115 to reach the detector 117. The resulting Raman spectral data of the sample is received by the detector 117, and the signal processing and / or digitization of the received Raman spectral data is processed by the electrical signal processor 113 (see, for example...). Figure 8 , 10 (13 and 16 to 17, which show exemplary Raman spectral data).

[0138] In some embodiments, the vesicle sample is scanned with an exposure time of 1 ms to 20 seconds. In some embodiments, the vesicle sample is scanned to capture both the bright Raman spectrum and the dark Raman spectrum of the sample.

[0139] In operation 706, controller 111 transmits Raman spectral data representing at least one Raman spectrum associated with one or more components of a sample of a vesicle to the stoichiometric model.

[0140] In some implementations, controller 111 determines the covariance matrix based at least on standardized Raman spectral data. Controller 111 determines the covariance matrix by generating a PxP symmetric matrix using at least one standardized Raman spectral data set. The covariance matrix can be a P-dimensional dataset, for example, a 2D dataset, or a 2048×2048 symmetric matrix with 2048 variables.

[0141] Controller 111 determines one or more eigenvectors and one or more eigenvalues ​​based on the covariance matrix. The eigenvectors and eigenvalues ​​are used to determine one or more principal component values. Controller 111 determines one or more eigenvalues ​​associated with the covariance matrix based at least on the orientation of the axis at the point of maximum variance in the Raman spectral data. The eigenvectors are commonly referred to as principal component values. Controller 111 determines one or more eigenvalues, where the eigenvalues ​​are coefficients associated with the eigenvectors, and the one or more eigenvalues ​​can provide the amount of variance carried in each principal component. Controller 111 organizes the eigenvectors from highest to lowest based at least on one or more eigenvalues, resulting in the principal component values ​​being sorted according to their importance. In some implementations, controller 111 determines the percentage of variance by dividing each eigenvalue by the sum of the eigenvalues.

[0142] Controller 111 generates a diagram representing the chemostometric model based on one or more principal component values. Controller 111 determines the eigenvectors of the chemostometric model by generating a matrix with columns of eigenvectors as defined above. Based at least on the eigenvectors, data associated with multiple Raman spectra are reoriented from the original axes to axes represented by one or more principal component values. Therefore, the PCA model is determined for one or more parameter levels associated with data representing one or more components of the training data (see, for example...). Figure 9 (This demonstrates the performance of an exemplary PCA model).

[0143] In some implementations, controller 111 generates a representation of the chemostometric model based on one or more principal component values. Controller 111 determines the eigenvectors of the chemostometric model by generating a matrix with columns of eigenvectors as defined above. Based at least on the eigenvectors, data associated with at least one Raman spectrum is redirected from the original axes to axes represented by one or more principal component values. Controller 111 uses the principal component values ​​to predict vectors and full rank through one or more regressions. Thus, a PLS model (see, for example) is determined for one or more parameter levels associated with data representing one or more components of a sample representing vesicles. Figure 9 and Figure 11 It demonstrates an exemplary PLS model.

[0144] In operation 708, controller 111 determines the levels of one or more parameters of one or more components of a vesicle sample based on a stoichiometric model.

[0145] Controller 111 determines one or more parameters representing the quality of one or more components of a sample representing vesicles. The one or more parameters representing the quality of one or more components indicate the identity of one or more components. In some embodiments, the one or more components include one or more solvents, one or more lipids, one or more surfactants, one or more LNPs, one or more micelles, or one or more intermediate phases, as discussed in detail above.

[0146] In some embodiments, controller 111 determines one or more parameters representing the quantity of one or more components of a sample representing vesicles. One or more parameters representing the quality of one or more components indicate the quantity of one or more components. In some embodiments, the quantity of one or more components includes concentration values ​​or ratios, as discussed in detail above.

[0147] In some aspects, process 700 may further include: one or more stoichiometric models comprising any one or more of the following: a PLS model, a PCA model, a PCR model, a minimum absolute shrinkage model, a LASSO model, an elastic network regression model, an SVM model, a neural network model, or any combination thereof. In some embodiments, the stoichiometric model comprises a PCA model and any one or more of the following: a PLS model, a PCR model, a minimum absolute shrinkage model, a LASSO model, an elastic network regression model, an SVM model, or any combination thereof.

[0148] In some respects, operation 704 may also include: containers such as reactors, vessels, sealed containers configured for storage, sealed containers configured for transport, etc.

[0149] In some aspects, operation 704 may also include: directly scanning lipid vesicles using the spectroscopic system 110. The controller 111 analyzes the Raman spectral data of a sample of vesicles and identifies one or more components of the sample of vesicles.

[0150] In some aspects, operation 704 may further include: controller 111 normalizing Raman spectral data representing at least one Raman spectrum associated with one or more components of a sample of a vesicle by applying one or more preprocessing operations to the Raman spectral data. In some embodiments, the one or more preprocessing operations include region selection, spectral averaging, convolution filtering, 一 First derivative, 二 The controller 111 applies region selection preprocessing to Raman spectral data of one or more components of a sample representing vesicles. In some embodiments, a portion of the entire Raman spectral data is used during region selection. Using only a portion of the entire Raman spectral data may have advantages in certain respects. For example, very high and very low wavenumber regions of Raman spectral data often have very low signal-to-noise ratios (SNR), which may present limited relevance and may lead to overfitting of the data when trained on noisy data. In another instance, different substances may be distinguished based on different regions of Raman spectral data in which specific peaks (i.e., wavenumbers) can be observed. In such respects, the remainder of the Raman spectral data may be observed to be less relevant, and data representing the remainder of the Raman spectral data may be discarded. In some embodiments, region selection is a hyperparameter. The controller 111 applies at least a second preprocessing operation that includes standard normal variables (SNVs). During SNV scaling, each spectral data point is scaled with a standard normal transformation as shown in equation (2) below: (2) Where, x i It is the first in the Raman spectral data i There are _ data points, where µ is the average intensity of the spectrum, σ is the standard deviation of the intensity, and x_i is the average intensity of the spectrum. iSNV It is x i The correction value.

[0151] In some respects, operation 708 may also include: any one of one or more parameters including concentration, fluorescence, refractive index, mass spectrometry, electrochemical behavior, or a combination thereof. As used herein, the “level” of a parameter is a qualitative or quantitative value corresponding to the parameter. For example, a parameter may be a concentration, where the level is a concentration value, such as in grams per milliliter. As another example, a parameter may be a concentration, where the level is selected from qualitative values, such as low, medium, and high.

[0152] This disclosure illustrates a process for monitoring the production of lipid vesicles. In some embodiments, the process for monitoring the production of lipid vesicles includes combining at least one lipid with at least one solvent to form a sample, and monitoring the sample using a spectroscopic system 110, as referenced above. Figure 6 and Figure 7 As described.

[0153] In some embodiments, the process of monitoring the production of lipid vesicles includes monitoring the presence or absence of one or more components of the lipid vesicles. These components include solvents, lipids, surfactants, lipid nanoparticles (LNPs), micelles, or intermediate phases of lipid nanoparticles (LNPs).

[0154] In some embodiments, the process of monitoring the production of lipid vesicles includes using a spectroscopic system 110 to monitor the quantity of one or more components of the lipid vesicles, as referenced above. Figure 6 and Figure 7 As described.

[0155] Experimental data

[0156] A. Exemplary stoichiometric model

[0157] In some implementations, the experimental data described below in this section includes at least the data mentioned above. Figures 6 to 7 Any of the operational methods described herein. The performance of the exemplary PCA model is... Figure 9 As shown in the example, the performance of the exemplary PLS model is determined by... Figure 11 , Figure 14 and Figure 15 Demonstration. Exemplary PCA and PLS models and references. Figures 6 to 7 The same model is used. The model is trained using compositional training data, which includes exemplary Raman spectral data (which may be actual data) and one or more exemplary known levels of parameters of one or more components of the vesicle.

[0158] Figure 8 This is a diagram 800 representing Raman spectral data for at least one Raman spectrum associated with one or more components of a vesicle. As shown in diagram 800, Raman spectral data for cholesterol 802, isopropyl myristate 804, stearate 806, and ethanol 808 are provided.

[0159] Figure 9 This is a figure 900 illustrating the performance of an exemplary stoichiometric model that uses principal component analysis to determine one or more parameters of the quality of one or more components of a sample representing vesicles. Specifically, Figure 9The identities of cholesterol 902, isopropyl myristate 904, stearate 906, and ethanol 908 were revealed.

[0160] Figure 10 This is a figure 1000 showing Raman spectral data for the quantification of single-component lipids in ethanol solution.

[0161] Figure 11 This is a diagram 1100 illustrating the performance of an exemplary stoichiometric model that uses PLS to determine one or more parameters representing the amount of isopropyl myristate in ethanol. (See diagram 1100) Figure 11 As shown, the predictive correlation coefficient (R²) for cross-validation was 0.999, and the RMSEC value was 0.358 mg / mL. Additionally, the RMSECV value was 0.575 mg / mL.

[0162] Figure 12 A figure 1200 shows Raman spectral data representing a comparison between the concentration of isopropyl myristate and the concentration of cholesterol.

[0163] Figure 13 Figure 1300 illustrates Raman spectral data representing at least one Raman spectrum associated with one or more components of a vesicle. Figure 1300 includes Raman spectral data for waterborne contamination.

[0164] Figure 14 This is a diagram 1400 illustrating the performance of an exemplary stoichiometric model using PLS to determine one or more parameters representing the amounts of isopropyl myristate and cholesterol in a mixture with ethanol. Figure 14 As shown in the figure, for cholesterol, the correlation coefficient (R2) is 1.000, 1.000, and the RMSECV is 0.308 mg / mL.

[0165] Figure 15 This is a diagram 1500 illustrating the performance of an exemplary stoichiometric model using PLS to determine one or more parameters representing the amounts of isopropyl myristate and cholesterol in a mixture with ethanol. Figure 15 As shown, for isopropyl myristate, the correlation coefficient (R2) was 1.000, 1.000, and the RMSECV was 0.671 mg / mL.

[0166] Figure 16 Figure 1600 shows the Raman spectral data from Figure 1300, where the red bar area 1601 represents data excluded from the quantitative model.

[0167] Figure 17This is a diagram 1700 of Raman spectral data obtained using the PLS model for the quantification of waterborne pollution (1701).

[0168] Figure 18 This is a diagram 1800 illustrating the performance of an exemplary stoichiometric model used to determine one or more parameters representing the amount of water, and more specifically, the quantification of isopropyl myristate and cholesterol in a binary mixture of ethanol. The concentrations of isopropyl myristate and cholesterol used to create the training model were calculated using a uniform design. The PLS algorithm was used to develop the quantitative model. Figure 18 As shown, the cross-validation correlation coefficient (R²) is 1.000, and the REMSEC value is 0.56 % v / v. The REMSECV value is also 0.56948 mg / mL. Therefore, based on the stoichiometric model, samples containing vesicles with aqueous contamination can be quantified.

[0169] B. Chemometric modeling data

[0170] To train the stoichiometric model described above, data representing samples containing one or more components were collected, as described in this paper, with some additional details described below.

[0171] To identify cholesterol, isopropyl myristate, and stearate, each component was dissolved in ethanol at a concentration of 5 mg / mL, and Raman spectra were collected. The PCA algorithm was applied to the collected Raman dataset to classify each component in PCA space. Scores on each principal component will allow for both identification and quantification.

[0172] To quantify single-component lipids, isopropyl myristate was mixed with ethanol at concentrations of 0, 6.5, 12.5, 25, 33.33, and 50% v / v. Raman spectra were collected, and the datasets were fed into a PLS algorithm to create a PLS quantitative model.

[0173] For the multi-component quantification of a mixture of isopropyl myristate and cholesterol, a uniform design was used to cover the concentration levels of interest. The methods described herein do not require the designed experimental procedures and can use randomly selected concentrations. The scope of the designed experiments is indicated in Table 1 below.

[0174]

[0175] Table 1

[0176] C. Pretreatment and Chemometric Model Construction

[0177] As discussed in this paper, the chemosiometric models were constructed using partial least squares (PLS) and / or principal component analysis (PCA) models. Both PLS and PCA models were modeled using MATLAB, developed by MathWorks.

[0178] Terms and Conditions

[0179] Implementation of this disclosure is disclosed in the following terms: Clause 1. A computer-implemented method in an analytical instrument support device, the method comprising: Raman spectral data associated with compositional training data is received by one or more processors, wherein the compositional training data includes data representing one or more components of vesicles; One or more preprocessing operations are applied by one or more processors to normalize the Raman spectral data; The covariance matrix is ​​determined by one or more processors based at least on standardized Raman spectral data; One or more processors determine one or more principal component values ​​associated with the covariance matrix; and One or more processors determine a stoichiometric model for one or more parameter levels associated with the data representing the one or more components of the vesicle, based on the one or more principal component values.

[0180] Clause 2. The computer-implemented method according to Clause 1, wherein: The one or more parameter levels represent the quality of the one or more components of the vesicle, wherein one of the one or more levels indicates the identity of the one or more components of the vesicle as modeled by the stoichiometric model; or The one or more parameter levels represent the quantity of the one or more components of the vesicle, wherein one of the one or more levels indicates the quantity of the one or more components of the vesicle as modeled by the stoichiometric model.

[0181] Clause 3. The computer-implemented method according to Clause 1 or Clause 2, wherein the one or more components of the vesicle comprise a solvent, lipid, surfactant, lipid nanoparticles (LNP), micelles, or an intermediate phase of LNP.

[0182] Clause 4. The computer-implemented method according to any one of Clauses 1 to 3, wherein one of the one or more components of the vesicle is a solvent comprising ethanol.

[0183] Clause 5. The computer-implemented method according to any one of Clauses 1 to 4, wherein one of the one or more components of the vesicle is a lipid comprising one or more of cholesterol, isopropyl myristate, or stearate.

[0184] Clause 6. The computer-implemented method according to any one of Clauses 1 to 5, wherein one of the one or more components of the vesicle is an LNP comprising isopropyl myristate and cholesterol.

[0185] Clause 7. A computer-implemented method according to any one of Clauses 1 to 6, wherein the one or more components of the vesicle comprise a solvent and lipids, and the composition training data comprises an experimental design for forming LNPs in an objective and controlled environment with concentrations of the lipids in the solvent.

[0186] Clause 8. The computer-implemented method according to any one of Clauses 1 to 7, wherein the experimental design is a uniform design.

[0187] Clause 9. The computer-implemented method according to any one of Clauses 1 to 8, wherein the chemometric model is a partial least squares regression (PLS) model and / or a principal component analysis (PCA) model.

[0188] Clause 10. A computer-implemented method on an analytical instrument support device, the method comprising: A chemostometric model for one or more parameter levels associated with composition training data is obtained by one or more processors, wherein the composition training data includes data representing one or more components of vesicles; Raman spectral data representing at least one Raman spectrum associated with one or more components of a sample of a vesicle are received by one or more processors; One or more processors transmit the Raman spectral data, representing at least one Raman spectrum associated with one or more components of the sample of the vesicle, to the stoichiometric model; and One or more processors determine the levels of one or more parameters of one or more components of the vesicle sample based on the stoichiometric model.

[0189] Clause 11. The computer-implemented method according to Clause 10, the method further comprising: The sample of the vesicles was scanned with light guided from a Raman spectrometer; In response to scanning the sample of the vesicle with the light, scattered light guided back from the sample of the vesicle is received via a detector; and In response to receiving the scattered light via the detector, a set of program instructions is generated to transmit the scattered light to an electrical signal processor by one or more processors. The electrical signal processor generates Raman spectral data representing the at least one Raman spectrum.

[0190] Clause 12. The computer-implemented method according to Clause 10 or 11, wherein the stoichiometric model is provided by a method comprising: The compositional training data representing the one or more components of the vesicles is received by one or more processors; and One or more processors determine the chemometric model for one or more parameter levels associated with the composition training data.

[0191] Clause 13. A computer-implemented method according to any one of Clauses 10 to 12, wherein

[0192] The level of one or more parameters represents the quality of one or more components, wherein the level of one of the one or more parameters indicates the identity of the one or more components as determined by the stoichiometric model; or

[0193] The level of one or more parameters represents the quantity of one or more components, wherein the level of one of the one or more parameters indicates the quantity of one or more components determined by the stoichiometric model.

[0194] Clause 14. The computer-implemented method according to any one of Clauses 10 to 13, wherein one or more components comprise a solvent, lipid, surfactant, lipid nanoparticles (LNP), micelles, or an intermediate phase of LNP.

[0195] Clause 15. One or more non-transitory computer-readable media having instructions stored thereon, which, when executed by one or more processing devices of an analytical instrument support device, cause the analytical instrument support device to perform the computer-implemented method according to Clause 1 or Clause 10.

[0196] Clause 16. An analytical instrument support system comprising: One or more processors; One or more non-transitory computer-readable storage media; and Program instructions stored on at least one of the one or more non-transitory computer-readable storage media for execution by at least one of the one or more processors, the program instructions including steps for implementing the method according to claim 1 or claim 10.

[0197] Clause 17. The analytical instrument support system according to Clause 16, wherein the program instructions are executed on a common computing device including at least one of the one or more processors.

[0198] Clause 18. An analytical instrument support device according to any one of Clauses 16 or 17, wherein the program instructions are executed on a computing device including at least one of the one or more processors, and wherein the computing device is located remotely from the analytical instrument associated with the analytical instrument support system.

[0199] Clause 19. An analytical instrument support system according to any one of Clauses 16 to 18. The program instructions thereon are executed on a user computing device including at least one of the one or more processors.

[0200] Clause 20. An analytical instrument support system according to any one of Clauses 16 to 19, wherein at least one of the one or more processors is disposed in an analytical instrument associated with the analytical instrument support system, and wherein the program instructions are executed on the at least one of the one or more processors.

[0201] Clause 21. An analytical instrument comprising: A light source, configured to direct light onto the surface of the sample; A spectrometer configured to acquire Raman spectra from the surface of the sample in response to the light source directing light onto the surface of the sample; One or more processors; One or more non-transitory computer-readable storage media; and A program instruction stored on at least one of the one or more non-transitory computer-readable storage media for execution by at least one of the one or more processors, wherein the program instruction, when executed by at least one of the one or more processors, causes the analytical instrument to perform the method according to claim 1 or claim 10.

[0202] Clause 22. A method for monitoring the production of lipid vesicles, the method comprising combining at least one lipid with a solvent to form a sample, and monitoring the sample using an analytical instrument according to Clause 21.

[0203] Clause 23. The method according to Clause 22, wherein the sample is monitored for the presence or absence of one or more components of the lipid vesicle.

[0204] Clause 24. The method according to Clause 22, wherein the sample is monitored for the quantity of one or more components of the lipid vesicles.

Claims

1. A computer-implemented method on an analytical instrument support device, the method comprising: Raman spectral data associated with compositional training data is received by one or more processors, wherein the compositional training data includes data representing one or more components of vesicles; One or more preprocessing operations are applied by one or more processors to normalize the Raman spectral data; The covariance matrix is ​​determined by one or more processors based at least on standardized Raman spectral data; One or more processors determine one or more principal component values ​​associated with the covariance matrix; as well as One or more processors determine a stoichiometric model for one or more parameter levels associated with the data representing the one or more components of the vesicle, based on the one or more principal component values.

2. The computer-implemented method according to claim 1, wherein... The one or more parameter levels represent the quality of the one or more components of the vesicle, wherein one of the one or more levels indicates the identity of the one or more components of the vesicle as modeled by the stoichiometric model; or The one or more parameter levels represent the quantity of the one or more components of the vesicle, wherein one of the one or more levels indicates the quantity of the one or more components of the vesicle as modeled by the stoichiometric model.

3. The method of claim 1, wherein the one or more components of the vesicle comprise a solvent, lipid, surfactant, lipid nanoparticles (LNP), micelles, or an intermediate phase of LNP.

4. The method of claim 3, wherein one of the one or more components of the vesicle is a solvent comprising ethanol.

5. The method of claim 3, wherein one of the one or more components of the vesicle is a lipid comprising one or more of cholesterol, isopropyl myristate, or stearate.

6. The method of claim 3, wherein one of the one or more components of the vesicle is an LNP comprising isopropyl myristate and cholesterol.

7. The method of claim 3, wherein the one or more components of the vesicle comprise a solvent and lipids, and the composition training data comprises an experimental design for forming LNPs in an objective and controlled environment with a concentration of the lipids in the solvent.

8. The method according to claim 7, wherein the experimental design is a uniform design.

9. The method according to claim 1, wherein the chemometric model is a partial least squares regression (PLS) model and / or principal component analysis (PCA).

10. A computer-implemented method on an analytical instrument support device, the method comprising: A chemostometric model for one or more parameter levels associated with composition training data is obtained by one or more processors, wherein the composition training data includes data representing one or more components of vesicles; Raman spectral data representing at least one Raman spectrum associated with one or more components of a sample of a vesicle are received by one or more processors; One or more processors transmit the Raman spectral data, representing at least one Raman spectrum associated with one or more components of the sample of the vesicle, to the chemometric model; as well as One or more processors determine the levels of one or more parameters of one or more components of the vesicle sample based on the stoichiometric model.

11. The method according to claim 10, further comprising: The sample of the vesicles was scanned with light guided from a Raman spectrometer; In response to scanning the sample of the vesicle with the light, scattered light guided back from the sample of the vesicle is received via a detector; as well as In response to receiving the scattered light via the detector, a set of program instructions is generated to transmit the scattered light to an electrical signal processor by one or more processors. The electrical signal processor generates Raman spectral data representing the at least one Raman spectrum.

12. The method of claim 10, wherein the stoichiometric model is obtained by a method comprising: The compositional training data representing the one or more components of the vesicles is received by one or more processors; and One or more processors determine the chemostometry model for the one or more parameter levels associated with the composition training data.

13. The method of claim 10, wherein The level of one or more parameters represents the quality of one or more components, wherein the level of one of the one or more parameters indicates the identity of the one or more components as determined by the stoichiometric model; or The level of one or more parameters represents the quantity of one or more components, wherein the level of one of the one or more parameters indicates the quantity of one or more components determined by the stoichiometric model.

14. The method of claim 13, wherein one or more components comprise a solvent, lipid, surfactant, lipid nanoparticles (LNP), micelles, or an intermediate phase of LNP.

15. One or more non-transitory computer-readable media having instructions stored thereon, which, when executed by one or more processing means of an analytical instrument support device, cause the analytical instrument support device to perform the computer-implemented method according to claim 1 or claim 10.

16. An analytical instrument support system, comprising: One or more processors; One or more non-transitory computer-readable storage media; as well as Program instructions stored on at least one of the one or more non-transitory computer-readable storage media for execution by at least one of the one or more processors, the program instructions including steps for implementing the method according to claim 1 or claim 10.

17. The analytical instrument support system of claim 16, wherein the program instructions are executed on a common computing device including at least one of the one or more processors.

18. The analytical instrument support system according to any one of claims 16 or 17, wherein the program instructions are executed on a computing device including at least one of the one or more processors, and wherein the computing device is located remotely from the analytical instrument associated with the analytical instrument support system.

19. The analytical instrument support system according to any one of claims 16 to 18, wherein the program instructions are executed on a user computing device including at least one of the one or more processors.

20. The analytical instrument support system according to any one of claims 16 to 19, wherein at least one of the one or more processors is disposed in an analytical instrument associated with the analytical instrument support system, and wherein the program instructions are executed on the at least one of the one or more processors.

21. An analytical instrument comprising: A light source, configured to direct light onto the surface of the sample; A spectrometer configured to acquire Raman spectra from the surface of the sample in response to the light source directing light onto the surface of the sample; One or more processors; One or more non-transitory computer-readable storage media; as well as A program instruction stored on at least one of the one or more non-transitory computer-readable storage media for execution by at least one of the one or more processors, wherein the program instruction, when executed by at least one of the one or more processors, causes the analytical instrument to perform the method according to claim 1 or claim 10.

22. A method for monitoring the production of lipid vesicles, the method comprising combining at least one lipid with a solvent to form a sample, and monitoring the sample using an analytical instrument according to claim 21.

23. The method of claim 22, wherein the sample is monitored for the presence or absence of one or more components of the lipid vesicle.

24. The method of claim 22, wherein the sample is monitored for the quantity of one or more components of the lipid vesicles.

Images