Enriched peptide detection by single molecule sequencing

Abstract

The inventions herein, e.g., relate to quantitative measurement of proteins, and provide significant improvements, e.g., in the sensitivity, accuracy, throughput and cost of measuring clinically important proteins in biological samples such as blood. The inventions herein, e.g., also relate to peptide library preparation for quantitative single molecule analysis.

Description

[0001] ENRICHED PEPTIDE DETECTION BY SINGLE MOLECULE SEQUENCING

[0002] 1 BACKGROUND

[0003] The entire content of each patent document and publication referenced in this application, including but not limited to those listed below, is hereby incorporated by reference herein in its entirety.

[0004] 1.1 BACKGROUND PATENTS

[0005] U.S. Provisional Patent Application No. 63 / 284,990, filed 12 / 1 / 21

[0006] U.S. Provisional Patent Application No. 63 / 288,987, filed

[0007] U.S. Provisional Patent Application No. 63 / 296,196, filed 1 / 4 / 22

[0008] U.S. Provisional Patent Application No. 63 / 303,417, filed 1 / 26 / 22

[0009] U.S. Provisional Patent Application No. 63 / 313,760, filed 2 / 25 / 22

[0010] U.S. Provisional Patent Application No. 63 / 348,213, filed 6 / 2 / 22

[0011] U.S. Provisional Patent Application No. 63 / 352,925, filed 6 / 16 / 22

[0012] U.S. Provisional Patent Application No. 63 / 373,875, filed 8 / 30 / 22

[0013] U.S. Provisional Patent Application No. 63 / 381,722, filed 10 / 31 / 22

[0014] US Patent No. 7,632,686 (application no. 10 / 676,005, entitled High Sensitivity Quantitation of Peptides by Mass Spectrometry filed 2 October 2003)

[0015] International Application No. PCT / US11 / 028569 (entitled Improved Mass Spectrometric Assays for Peptides filed 15 March 2011)

[0016] International Application No. PCT / US 13 / 48384 (entitled Multipurpose Mass Spectrometric Assay Panels for Peptides)

[0017] International Application No. PCT / US 12 / 042,931 (entitled Magnetic Bead Trap and Mass Spectrometer Interface)

[0018] 1.2 SINGLE MOLECULE SEQUENCING USING NANOPORES : http s : / / nanoporetech . com / PCT / GB2020 / 053082 US11098355 US11168363

[0019] US20200239950A1

[0020] US 2017021955

[0021] US010814298

[0022] US20210147904

[0023] 1.3 REVERSE TRANSLATION OF PEPTIDE SEQUENCE TO DNA (OR RNA, ETC.)

[0024] FOLLOWED BY DETECTION USING HIGH-THROUGHPUT NUCLEIC ACID

[0025] SEQUENCING PLATFORMS:

[0026] “Proteocode” technology developed by Encodia: https: / / www.encodia.com / technology).

[0027] US20180201980A1

[0028] US20180328936A1

[0029] US20200348308A1

[0030] US20210254047A1

[0031] US20210302431A1

[0032] WO2017192633A1

[0033] “ProtSeq” technology developed by Google:

[0034] US 2021 / 0102248

[0035] US 2021 / 0079557

[0036] US 2021 / 0079398

[0037] US 2021 / 0171937

[0038] 1.4 AFFINITY REAGENT IMAGING PLATFORMS

[0039] Nautilus: https: / / www.nautilus.bio

[0040] US 2020 / 0318101

[0041] US 2021 / 0358563

[0042] US 2021 / 0239705

[0043] US 2021 / 0101930

[0044] US 2020 / 0082914

[0045] US 10,948,488 B2 PEPTIDE DEGRADATION WITH OPTICAL DETECTION OF TERMINAL AMINO

[0046] ACIDS

[0047] Quantum-Si: https: / / www.quantum-si.com / products-and-technology / and the following patent filings:

[0048] US11175227

[0049] US20160041095A1

[0050] US20200123593A1

[0051] US20200123594A1

[0052] US20200395099A1

[0053] US20210121875A1

[0054] US20210139973A1

[0055] US20210217800A1

[0056] US20210270740A1

[0057] US20210331170A1

[0058] US20210354134A1

[0059] WO2021086945A1

[0060] WO2021086954A1

[0061] WO2021146475A1

[0062] WO2021216763A1 FLUOROSEQUENCING AND FRET FINGERPRINTING:

[0063] US 9,625,469

[0064] US 10,545,153

[0065] US 2021 / 032536

[0066] US2018 / 0201980

[0067] US2018 / 0201980

[0068] US2021 / 03024

[0069] US2021 / 0254047

[0070] US20150087526A1

[0071] US20180328936A1

[0072] US20200018768A1

[0073] US20200123593A1 US20200123594A1

[0074] US20200124613A1

[0075] US20200231956A1

[0076] US20200348308A1

[0077] US20200400677A1

[0078] US20210221839A1

[0079] US20210331170A1

[0080] WO2016164530A1

[0081] WO2017192633A1.

[0082] WO2019222527A1

[0083] W02020014586A9

[0084] W02021086908A1

[0085] WO2021216763A1

[0086] WO2021111125A1

[0087] 2 FIELD OF THE INVENTION:

[0088] 2.1 PROTEIN QUANTITATION.

[0089] The inventions herein relate to quantitative measurement of proteins, and provides significant improvements in the sensitivity, accuracy, throughput and cost of measuring clinically important proteins in biological samples such as blood. More than 100 different proteins are currently measured by clinical diagnostic tests in blood (1), each requiring a separate test and a separate aliquot of sample. Such tests are typically immunoassays, and make use of indirect detection of protein targets by antibodies, opening the door to a variety of interferences and associated clinical errors (2). The cost and complexity of this paradigm for clinical laboratory testing severely constrains the health benefits obtainable from measurement of clinical biomarker proteins, and effectively precludes emerging applications such as high frequency longitudinal testing to establish personal biomarker baselines and health models. The inventions herein also relate to peptide library preparation for quantitative single molecule analysis.

[0090] 2.2 EARLIER AFFINITY ENRICHMENT METHODS: SISCAP A-MS.

[0091] In the past, significant progress was made in improving the specificity, multiplexability and sensitivity of protein tests through the introduction of quantitative mass spectrometric protein assays, particularly those using specific proteolytic peptides as quantitative surrogates for their parent proteins and enriching those peptides using peptide-specific enrichment reagents such as anti-peptide antibodies (e.g., the SISCAPA technology, (3, 4)). These advances have improved patient care (e.g., the SISCAPA test for the thyroid cancer marker thyroglobulin performed by leading clinical reference labs in the US and Canada (5, 6), and the recently introduced SISCAPA assay for SARS-CoV-2 NCAP protein (7)), and are widely used in pharmaceutical research and development to measure biomarkers, therapeutic proteins and drug targets with high precision.

[0092] However, the requirement for a mass spectrometer as the final detector in such assays represents a significant barrier to adoption due to capital cost (typically -$500,000), operator expertise required, limited throughput (typically 2-15 minutes per sample), and unsuitability for ultimate point-of-care use. In addition, the practical sensitivity of mass spectrometers for peptide detection is limited, with the best current instruments requiring at least 10 amol (-6 million molecules) of a peptide for reliable detection, far more than would be required in principle if individual molecules could be counted reliably. It is an object of the present invention to overcome these barriers by enabling the use of single molecule detection techniques in quantitative protein assays.

[0093] 2.3 SINGLE MOLECULE METHODS.

[0094] Several technologies are being advanced for sequence-sensitive single molecule peptide characterization and detection using concepts and methods initially developed for DNA and RNA sequencing. These include nanopore sequencing, “reverse translation” of peptide to DNA sequences, cyclical degradation with electronic detection of terminal amino acids, optical detection of single molecule epitopes, etc., as described below. Nucleic acid versions of such methods typically aim to provide enormous throughput (e.g., gigabases per run) of sequence (i.e., digital) data - a feature required to address whole genome, whole exome, or RNASeq sequencing requirements - but are not focused on precise quantitative (i.e., analog) measurements of the amount of a particular type of molecule. Publications in these field rarely (if ever) use statistical terms related to quantitative precision (such as variance, coefficient of variation (CV), or accuracy): in contrast, these terms are commonly used in discussions of protein quantitation methods where diagnostic accuracy often requires preset precision (e.g., a CV of 5%).

[0095] Several significant barriers impede the use of these technologies for the analysis of peptides and proteins. Peptides and proteins are made of 20 common amino acids, as opposed to only 4 common bases in either RNA or DNA, requiring a much greater degree of analytical discrimination to sequence peptides as opposed to DNA. An additional consequence of the greater variety of amino acids, compared to bases, is that peptides and proteins have a very wide variety of physical properties (e.g., number, polarity and localization of electric charges, solubility, chemical reactivity, inter-molecular interactions, etc.) as compared to most nucleic acids, which are uniformly negatively charged along their lengths, with generally unreactive bases attached. In addition, there is no biological equivalent of the polymerase chain reaction (PCR) for peptides and proteins, or of any enzymatic process capable of copying a protein molecule directly, thus eliminating many of the most powerful methods used in nucleic acid preparation for sequencing. Perhaps most importantly, without PCR or an equivalent amplification method for peptides, these methods have limited dynamic range. Genomic DNA has a very limited dynamic range (genes are generally present at near equal stoichiometry). While RNA or cell-free DNA can be present at widely varying abundances, PCR can be used to amplify low abundance molecules. In contrast, important protein-containing samples such as blood, plasma or tissues, have documented dynamic ranges in excess of 1011(difference in molar amounts between the highest and lowest abundance proteins for which there is a clinical need to measure), with no practical method of amplifying the numbers of low abundance molecules. As a result, there are major gaps in the generality, specificity, and robustness of peptide methods derived from nucleic acid technologies, and thus significant barriers to their applicability to the problem of quantitative protein measurement.

[0096] 2.4 THE PRESENT INVENTIONS.

[0097] To overcome these limitations and enable use of a “sequence-sensitive single molecule detector” instead of a mass spectrometer to detect and measure peptide molecules, major challenges in the preparation of peptide samples and their presentation to such detectors must be addressed. The present invention provides a general approach to the preparation of peptide libraries for quantitative single molecule analysis, and specific implementations appropriate for use with several alternative single molecule detectors (nanopores, optical imaging systems, and single molecule stepwise sequencing systems).

[0098] A key obstacle in formulating the invention has been the multi-dimensional nature of the problem, encompassing as it does the areas of protein and peptide chemistry, oligonucleotide chemistry and sequence design, antibody selection, single molecule detection by optical, chemical and electrical technologies, and requirements of specific clinical diagnostic assays. Key aspects of the invention involve adaptation of technologies from each of these areas in a novel combination.

[0099] The invention preserves the fundamental benefits of direct detection of analyte molecules (a strength of mass spectrometry in comparison with indirect detection methods such as immunoassays), while offering the potential for improved test sensitivity, sequence specificity, and lower cost - all of which improve commercial competitiveness against legacy immunoassay technologies and enable expanded use of protein biomarkers in medicine and pharmaceutical R&D. Using reagents and methods of the present invention, a substantial improvement in the throughput, cost and sensitivity of protein analysis can be achieved. In the case of nanopore sequencing, taken as an example of sequence-sensitive single molecule detection, we can, in principle, estimate the performance of a system optimized for peptide quantitation. Assuming that 1) peptides can be delivered as oligonucleotide constructs of about 50 bases in length; 2) nanopore sequencers process (read) oligonucleotides at a rate of approximately 400 bases / sec; 3) accurate measurement of the amount of a peptide requires detection and counting of less than 5,000 molecules; and 4) a commercially available nanopore cartridge contains 3,000 simultaneously readable nanopores, it would be theoretically possible to identify and precisely measure 30 peptide targets (representing 30 distinct clinically-relevant proteins) in a single sample in approximately 6 seconds. Using DNA barcoding methods described herein to multiplex samples, 96 samples could be analyzed in 10 minutes using a single cartridge, compared to approximately 10 hr using liquid chromatography - mass spectrometry (LC-MS): an advantage of 60-fold in speed with less than I / IO*11the equipment cost. Relatively inexpensive benchtop commercial devices exist capable of operating 48 such cartridges simultaneously, providing potential throughput for such a 30-plex biomarker panel of 4,608 samples in 10 min, or more than 600,000 samples per 24-hr day (assuming that sample preparation could keep up with this throughput!).

[0100] Using a completely different single molecule detection method based on optical imaging of molecules immobilized on a planar array, similar increases in analytical throughput can be obtained. An array capacity of 2,000,000,000 individual molecules would allow 400,000 peptide abundance measurements, assuming that 5,000 molecules need to be counted for each measurement and that all the different molecules can be brought to near equal stoichiometry. A single run of such a system, requiring approximately 1 day for optical readout, could therefore measure 40 proteins in 10,000 samples per day.

[0101] Furthermore, the ability to recognize and count individual analyte molecules can, at least in theory, offer the maximum assay sensitivity possible by any method, approximately 1,000 times as sensitive as mass spectrometry (-5,000 vs 6,00,000 molecules required for quantitative measurement, respectively). Such an improvement in sensitivity would enable precise measurement of almost all the 100+ clinically- established blood protein biomarkers in much less than 1 microliter (1 / 20111of a drop) of blood.

[0102] These advances in throughput and sensitivity can translate into significant improvements in the cost of measuring the current menu of approximately 115 blood protein biomarkers. Current clinical laboratory analyzers measure one protein at a time, typically by means of specific immunoassays, at an average cost of $5-10 per protein per sample, and requiring 50-100 uL of sample per measurement (which explains why 5-10 mL of blood is typically drawn by venipuncture when blood tests are required). A significant downside of this paradigm is the limitation it places on use of protein panels: measured one at a time, a panel of proteins would cost $5-10 times the number of proteins - a strong disincentive impeding the application of panels despite their greater diagnostic information content. Using single molecule methods, it is estimated that a single run costing $10,000 and yielding 400,000 protein measurements would lower the cost per individual clinical protein measurement to $0,025. This cost structure would revolutionize clinical diagnostics and enable major advances in disease detection and management.

[0103] However, this enormous advance in performance provided by sequence-sensitive single molecule detectors is only realizable if several very challenging problems can be solved that currently impede their use to measure a wide range of clinically-important proteins in diagnostic samples such as plasma and whole blood. These problems relate to the preparation of peptide libraries for quantitative single molecule analysis, and are successfully overcome by the present invention, which provides improvements in quantitation (by providing novel internal standards), sensitivity (by enriching low-abundance targets), dynamic range (by enabling stoichiometric flattening), and analytical workflow (by providing applicable chemistries for sample preparation).

[0104] 3 BACKGROUND OF THE INVENTION:

[0105] The present inventions address several major challenges in protein analysis, making use of a number of methods well-known in the art, in novel combinations and in combination with entirely novel concepts disclosed herein. 3.1 PROTEIN QUANTITATION CHALLENGES .

[0106] Quantitative measurements of protein biomarkers, drugs and drug targets are important in many areas of medical practice, pharmaceutical trials and biological research. The importance of improving such measurements in terms of sensitivity, specificity, and generality is nowhere more significant than in the context of blood, the primary clinical specimen. Blood represents the largest and deepest version of the human proteome present in any sample: in addition to the classical “plasma proteins” and cellular proteins of red cells, white cells and platelets, it contains all tissue proteins (as leakage markers) plus very numerous distinct immunoglobulin sequences (8). In addition to the large number of proteins present, proteins in plasma exhibit an extraordinary dynamic range in abundance: more than 10 orders of magnitude in concentration separate albumin and the rarest proteins now measured clinically. Abundant scientific evidence, from proteomics and other disciplines, suggests that among these are proteins whose abundances and structures change in ways indicative of many, if not most, human diseases. Nevertheless, only about 100 proteins are currently used in routine clinical diagnosis ( ), while the rate of introduction of new protein tests approved by the US FDA has paradoxically declined over the last two decades to about one or two new protein diagnostic markers approved per year. Furthermore, it appears that the clinical value of most such tests would be substantially improved if the results were interpreted in terms of patient- specific (i.e., personalized) baselines (rather than population reference intervals) - an advance that is currently inhibited by the cost and inconvenience of collecting a series of baseline samples from each patient before the emergence of major disease processes (9). Major advances in clinical diagnostics and pharmaceutical research are to be expected if certain technical problems in sample collection, preparation and analysis are solved.

[0107] Current methods of protein analysis, including those used in clinical laboratories for analysis of samples like blood, have limitations that significantly impact their utility, both in research and clinical practice. High-precision tests for clinical use are expensive, limited to a small menu of proteins, and require expensive equipment. Research methods, e.g., those of proteomics, measure many proteins, but with limited precision, low throughput and high cost. There is therefore a need for improvements in protein measurement. 3.2 PEPTIDES AS QUANTITATIVE SURROGATES FOR PROTEINS

[0108] In many applications aimed at protein quantitation, one can use a single peptide as a quantitative surrogate for the parent protein, provided that there is one (or some other known number) of copies of the peptide per protein molecule; i.e., that the peptide molar amount (or number of molecules) is equal to (or some known multiple of) the protein’s molar amount (or number of molecules). Using the known sequence of a target protein and / or experimental data, one can select one or more proteolytically-derived peptide segments within it as "target peptides" (herein referred to as TARGET(s)) to be measured as surrogates for their parent proteins. A good target peptide for quantitation purposes is one that is a) proteotypic for the protein (i.e., occurs in no other protein of the species from which the sample is derived); b) occurs a known number of times (usually once) in the protein sequence, allowing the peptide to be used as a surrogate measure of the molar amount of the protein; c) is efficiently detected by a chosen detector; and d) behaves reliably in a practical sample preparation workflow appropriate to the assay objectives (which may include, for example, specific binding and enrichment compared to other un-selected peptides). Methods for selection of TARGET peptides from a wide range of target proteins for conventional mass spectrometric detection is well-known in the art, but not directly relevant to selection of optimal peptides for single molecule detection. .

[0109] 3.3 PROTEOLYTIC DIGESTION

[0110] Digestion of proteins to peptides serves to “simplify” the structure of a protein sample, by eliminating complicated protein shapes (and their associated unique physical properties and protein: protein interactions), at the expense of increasing the numbers of molecules present. In other words, the immense variety of folded protein structures present in a biological sample is transformed by digestion into a larger set of essentially unstructured short, linear peptides. Proteins exhibit a very wide range of physical properties, ranging from soluble to insoluble, compact to extended, positively to negatively charged, with half-lives of seconds to months, and thus each protein represents an individual challenge in terms of handling and measurement. However, proteolytic digestion of a given protein to peptides generally yields a mixture of peptide molecules from which an example can almost always be chosen that is unique to a given target protein (and thus can serve as a quantitative surrogate for it) and has properties compatible with a selected measurement method (encapsulated by the aspirational phrase “in every bad protein there is at least one good peptide”). For this reason, peptide-level detection is less susceptible to interferences, and more compatible with universal sample preparation methods, than protein-level detection. A typical human protein yields about 50 peptides upon digestion with trypsin, and thus a sample containing, for example, 5,000 proteins is likely to yield a tryptic digest containing 250,000 different peptides. Peptides of the length of typical tryptic peptides (5 to 25 amino acids in a typical tryptic digest) do not generally exhibit stable folded structures and thus do not generally interact with one another to form stable multi- peptide structures. This overall absence of stable interactions between digest peptides overcomes the major source of interference and error in technologies such as conventional immunoassays.

[0111] Proteolytic digestion is widely used in proteomics to fragment proteins for analysis by mass spectrometry (10) and other analytical methods. Digestion of a sample such as plasma is typically carried out by first denaturing the sample proteins (e.g., with detergents such as deoxycholate, organic solvents, urea or guanidine HC1), reducing the disulfide bonds in the proteins (e.g., with tris(2-carboxyethyl)phosphine (TCEP), dithiothreitol or mercaptoethanol), alkylating the cysteines to prevent re-formation of disulfides (e.g., by addition of iodoacetamide which reacts with the free -SH group of cysteine), quenching excess iodoacetamide by addition of more dithiothreitol or mercaptoethanol, and finally (after removal or dilution of the denaturant) addition of the selected proteolytic enzyme (e.g. trypsin, Lys-C, etc.,), followed by incubation to allow digestion. Following incubation, the action of trypsin is terminated, either by addition of a chemical inhibitor (e.g., TLCK) or by denaturation (through heat or addition of denaturants, or both) or removal (if the trypsin is on a solid support) of the trypsin. Digestion destroys protein: protein interactions and thus generally eliminates interferences that occur in conventional immunoassays.

[0112] A very wide variety of proteolytic digestion protocols have been developed, and some have been shown to exhibit extremely high quantitative reproducibility when implemented on automated platforms (4). Most such protocols involve use of a single proteolytic step with a single enzyme (typically trypsin), while in a few cases two enzymes are used together (e.g., Lys-C and trypsin) in order to improve efficiency: Lys-C is smaller than trypsin and more stable at elevated temperature and in the presence of denaturants, and therefore able to cleave proteins that are otherwise relatively resistant to trypsin attack. In some cases this approach makes use of sequential digestion by Lys-C followed by trypsin, with these two steps carried out at different temperatures or at different denaturant concentrations. In contrast to embodiments described herein, the sequential use of Lys-C and trypsin to improve digestion efficiency does not allow oriented construction of peptide-polymer constructs as disclosed in the present invention.

[0113] 3.4 THE DYNAMIC RANGE PROBLEM

[0114] Quantitative detection of peptides by any method faces a major challenge in the form of the vast dynamic range of protein concentrations in samples of interest. In the main clinical specimen (blood serum or plasma) proteins of clinical interest span more than 10 orders of magnitude (10 billion-fold) between the highest abundance proteins (albumin in plasma, or hemoglobin in whole blood) and low abundance proteins of interest (e.g., thyroglobulin (Tg) in the blood of a thyroid cancer survivor). Thus calculations based on the known amounts of various proteins in human plasma show that for every molecule of a selected peptide unique to thyroglobulin in the tryptic digest of a given volume of plasma, there are ~40 billion molecules of other peptides. Comparing the selected thyroglobulin peptide abundance with that of a selected peptide from albumin, there are more than 400,000,000 copies of the albumin peptide for each molecule of the Tg peptide. An example panel of clinically important proteins is shown in Figure 1, comprising a high-abundance protein (transferrin), and lower abundance proteins soluble transferrin receptor (sTfR) and hepcidin, as well as thyroglobulin (present at 0.5 ng / ml - a level indicative of thyroid cancer recurrence in someone treated for that cancer). In a plasma digest, a transferrin peptide is expected to outnumber sTfR peptides by almost 1,000 to 1; to outnumber hepcidin peptides by almost 5,000 to 1, and to outnumber Tg peptides by 28,000,000 to 1. These proteins are measured today in separate assays, each optimized for a different abundance level, and typically offering an assay dynamic range of -1,000 (i.e., a range in which most clinical specimen concentrations of that protein are expected to fall).

[0115] Wider dynamic ranges have been achieved in detection systems reliant on amplification (e.g., PCR assays for nucleic acids and proximity ligation (11) or Somascan (12) assays for proteins). However, amplification-based assay systems sacrifice certainty regarding analyte identity, since the actual molecular targets are not themselves observed or measured by the ultimate detectors used in such assays (which detect binding reagents such as antibodies instead), with the result that unexpected interfering molecules can be measured and genuine analyte molecules can fail to generate signal. Confidence that the intended analyte, and only this analyte, is being measured requires direct analyte detection by a detector capable of discriminating the correct analyte from all others, as exemplified by sequence sensitive single molecule detectors used in the present invention.

[0116] 3.5 LIMITATIONS OF MASS SPECTROMETRY AND SISCAP A

[0117] Using direct analyte detection (whether by mass spectrometry or single molecule detection) with samples such as blood plasma comprising a very wide dynamic range of protein targets represents a major technical challenge. A practical technology for measuring the clinical plasma proteome should be capable of accurately quantitating panels of specific biomarker proteins spanning all abundance levels in a single aliquot of sample. To date, the favored approach to this problem has been a combination of differential enrichment of peptides during sample preparation (bringing target peptides into more equal ratios) with mass spectrometry detection (providing a close approximation to sequence-based analyte identification) - an approach termed SISCAPA. The SISCAPA method described in previous disclosures (US7632686) and publications (3, 9, 13-15) is a general approach for protein quantitation involving digesting proteins (e.g., with trypsin) into peptides that can be enriched by specific affinity capture and further fragmented in a mass spectrometer (e.g., by LC- MS / MS) to generate a sequence-based identification and a measure of amount by comparison to an internal standard. This approach combines the advantages of classical immunoassays (sensitivity, throughput) with those of mass spectrometry (specificity, multiplexability, wide linear dynamic range), while overcoming the limitations of each. Challenges remain, however, in sensitivity (mass spectrometry typically requires millions of molecules to generate a reasonably precise signal), throughput (MS-based assays typically require 5-30 minutes per sample), cost (mass spectrometers are expensive, e.g., $500,000, and require expert operators), and robustness (MS-based systems are typically confined to large institutions with sophisticated infrastructure).

[0118] A major improvement would result from the replacement of mass spectrometry by a sequence-sensitive single molecule detector, coupled with sample preparation technology capable of delivering to the detector the small numbers (e.g., 100-10,000) of purified analyte molecules that, when counted by the detector, generate the required assay precision (in this case determined by counting statistics).

[0119] No mature, practically-implementable technology currently exists that is capable of preparing peptide libraries from complex sample digests for single molecule analysis over such a wide dynamic range.

[0120] 3.6 PEPTIDE CHEMISTRY

[0121] A wide range of chemical modifications have been devised that can be of use in the preparation of peptides and digests for analytical applications of the invention. In some embodiments described herein a peptide is chemically modified, for example to create a linkage to another molecule during assembly of a novel multi-part construct. Site specific linkage chemistries are known for amino groups (e.g., the n-terminal amino group, and the epsilon amino group of lysine); for carboxyl groups (e.g., the c-terminal carboxyl, and sidechain carboxyls of aspartic and glutamic acids); sulfhydryl groups of cysteine residues; and a variety of other less frequently used chemistries. Of these, only the amino and carboxyl groups are available on almost all peptides, making them the preferred attachment points for general methods applicable to a wide variety of peptides. While many reagents have been identified that react preferentially with amino groups, the most common chemistries are n- hydroysuccinimide (NHS) esters and their more soluble sulfo-NHS derivatives. The n- terminal amino group and epsilon amino group of lysine have significantly different pK values, offering the potential to modify one in preference to the other, though this distinction is not absolute. Nevertheless, protocols have been devised (16) that couple NHS-derivative small molecules (e.g., TMT labels used in mass spectrometric detection) to peptide amino groups with very high efficiency (>95%) using very little excess reagent (less than 2-fold excess of reagent over amino groups), illustrating the feasibility of quantitative modification of amino groups in complex peptide mixtures such as proteolytic digests. Carboxyl groups can also be modified using chemistries involving carbodiimides (e.g., EDC: lethyl-3-(3- dimethylaminopropyl)carbodiimide); however the lack of site specificity of such reactions (between c-terminal and amino acid side chain carboxyls) restricts their usefulness for site- specific approaches. A variety of reagents exist that are capable of introducing “click” chemistry functional groups into peptides, often by reaction of an NHS-derivative of a click functionality with a peptide amino group, and oligonucleotides, often by incorporating an amino-derivatized base in a DNA sequence during synthesis and subsequently adding a click group by reaction with this amino group (17). Examples of effective click reagent pairs useful in creation of constructs according to the invention include i) reaction of an azide with an alkyne functionality (some requiring Cu(I) catalysis, which is less preferred in some embodiments); ii) reaction of an azide with a cyclooctyne such as DBCO (dibenzocyclooctyne, also called DIBO), Aza- dibenzocyclooctyne (ADIBO) or BCN (bicyclo[6.1.0]non-4-yne) by means of a strain- promoted alkyne cycloaddition (SPAAC) reaction without the need for a Cu catalyst; and iii) reaction of a tetrazine (Tz, such as methyltetrazine) with a trans-cyclooctene (TCO), also without the need for a Cu catalyst.

[0122] 3.7 INTERNAL STANDARDIZATION

[0123] Use of an internal standard in an analytical assay is highly desirable as it provides a stable reference against which the desired analyte can be measured. In the case of mass spectrometric detection of peptides, a synthetic stable isotope labeled version of a target peptide can easily be made and used as an internal standard (the well-known method of “isotope dilution mass spectrometry”). The approach works well because the labeled and unlabeled peptides are chemically and structurally identical, and thus behave the same through any sample preparation protocol, yet can be distinguished reliably by measuring their masses in the final mass spectrometer detection step. Since the labeled peptide is added at a known concentration, the ratio between the amounts of the natural and isotopically labeled forms detected by the final MS analysis allows the concentration of the natural peptide in the sample mixture to be calculated. The approach can be multiplexed to cover multiple peptides measured in parallel, and can be automated through computer control to afford a general system for protein measurement (13).

[0124] Single molecule detectors are unable to measure peptide mass accurately enough (or in most cases at all) to use stable isotope versions as internal standards in this manner. Hence there is a need for an alternative peptide labeling strategy to create single molecule internal standards capable of a) behaving like the targeted peptide analyte during the steps of sample preparation, while b) being clearly distinguishable from the target by the chosen single molecule detection technology. Use of the term “standard” in this specific sense is distinct from other forms of “standards” that can be introduced into workflows for quality control of separations, monitoring of chemical reaction yields, etc., rather than improving quantitation of a single specific analyte.

[0125] 3.8 PEPTIDE-SPECIFIC BINDERS

[0126] A variety of types of biologically derived antibodies (e.g., polyclonal, monoclonal and oligoclonal antibodies derived from mice, rabbits, humans, camelids and other species), molecules derived from antibodies by molecular biology techniques (e.g., antibodies selected from libraries using phage display and other techniques), aptamers (based on DNA, or RNA, and including a variety of modified bases and backbones), and other molecular constructs can be created that are capable of specifically binding a peptide (“BINDERs”). For example, a TARGET peptide can be coupled to a carrier protein (e.g., keyhole limpet hemocyanin: KLH) and used to immunize an animal (such as a rabbit, mouse, chicken, goat, camelid or sheep) by one of the known protocols that efficiently generate anti-peptide antibodies. Experience with the SISCAP A technology (3, 9, 13-15) has shown that antibodies, preferably monoclonal antibodies, can be developed that bind and capture a specific low abundance tryptic peptide from the digest of a very complex sample such as human blood plasma (which may contain 250,000 distinct peptides, some at very high abundance), and thereby enrich the peptide substantially (e.g., more than 10,000-fold). Discovery of such BINDERS (e.g., antibodies) requires use of very specific screening processes to find reagents that do not bind non- TARGET peptides and retain the TARGET peptide long enough to wash non-binding peptides away (typically 10-15 minutes in many automated protocols). The screening process does not assess equivalence of binding TARGET and STANDARD (stable isotope-labeled peptide) since it is known there will be no difference (at least for15N and13C isotopic labels).

[0127] If, however, a peptide TARGET and its cognate internal STANDARD molecule are not chemically identical (as is the case with stable isotope labeled standards), the very specificity of effective BINDERS creates a major problem: if a BINDER binds a STANDARD more or less tightly (or with different kinetics) than its cognate TARGET peptide, then binding will impact the ratio of TARGET to STANDARD molecules and lead to an incorrect assay result. The selection of TARGETS, STANDARDS and BINDERS that successfully preserve the quantitative ratio is therefore critical for enablement of quantitative internally-standardized single molecule detection.

[0128] 3.9 STRENGTHS AND LIMITATIONS OF SINGLE MOLECULE DETECTION

[0129] The ability to detect a single molecule of an analyte provides the maximum theoretically possible analytical sensitivity. While only a few current clinical assays are sensitivity-limited, sensitivity is nevertheless a major barrier for many emerging and potentially clinically-useful protein biomarkers. Single molecule sensitivity would also allow measurement of the 100+ proteins of the current clinical laboratory menu in much smaller samples than the 10-100uL plasma currently required, and therefore reduce the need for phlebotomy by enabling use of tiny samples such as dried blood microsamples. The few single molecule methodologies so far developed for practical application to protein analytes (e.g., Quanterix SIMOA technology: https: / / www.quanterix.com / simoa-technology / ) make use of indirect detection (e.g., using antibodies to identify analyte molecules) and are therefore subject to the well-known range of immunoassay interferences, non-linear responses, and limited multiplexability.

[0130] The ability to determine the partial or full sequence of a biomolecule confers a further major advantage: improved confidence in its identity. For biopolymers such as nucleic acids and proteins, sequence information can identify the analyte unambiguously, thus enabling direct analyte detection. Methods that make use of antibodies to recognize intact protein analytes (e.g., immunoassays) cannot provide this level of certainty, and are classified as indirect detection methods.

[0131] It is thus highly desirable in protein analysis to employ single-molecule methods that provide sequence information (18, 19) or structural information closely tied to sequence. A number of such methods are being developed, in most cases making use of technology foundations created for use in nucleic acid (e.g., DNA and RNA) analysis, where sequence information is the primary deliverable.

[0132] The present invention provides reagents, methods, and kits for the preparation of peptide libraries suitable for quantitative analysis by a variety of sequence-sensitive single molecule detection technologies, including, but not limited to, the following examples: 3.9.1 Single molecule sequencing using nanopores.

[0133] Biopolymers (nucleic acids and polypeptides) can pass through nanopores (both biological and inorganic) in suitable membranes. Signals (e.g., through-pore ion current, or cross-pore tunneling current) recorded during transit of the analyte molecule through the pore reflect differences in blockade of ions flowing through the pore by the side chains of the biopolymer (i.e., bases or amino acids) in the pore’s throat region. Nanopore methods of DNA and RNA sequencing have been developed and successfully commercialized {(20, 21)).

[0134] Nanopore analysis of peptides and proteins is advancing rapidly {(19, 22-25)), but discrimination of 20 different amino acids presents a far greater challenge than discrimination of 4 nucleic acid bases. The most mature approach for nanopore analysis of peptides is one involving in-line linkage of peptides and nucleic acids into a hybrid polymer, allowing use of some features of a successful commercially-available DNA sequencing platform to be applied to peptides (e.g., international publication WO 2021 / 111125 Al). Similar methods are likely to work with a variety of alternative platforms including, but not limited to, alternative biological nanopores (26, 27), inorganic nanopores (28, 29), DNA-origami nanopores (1,_2)_ and the like.

[0135] 3.9.2 Single molecule characterization using an Affinity Reagent Imaging Platform (ARIP) in which fluorescent BINDERS recognize molecular features.

[0136] Single protein molecules can also be arrayed in a regular pattern on a planar surface and probed by a succession of “promiscuous” binding agents to build up a pattern of epitope occurrences in each molecule (30-32). Machine learning approaches can be used to interpret these epitope occurrence patterns to identify most proteins produced in a given organism despite the stochastic nature of individual binding events. In the context of short peptides, rather than whole proteins, this approach does not deliver direct peptide sequence information.

[0137] 3.9.3 “FRET” fingerprinting of peptides moving through a protease

[0138] A limited but sequence-specific fingerprint of a peptide can be accomplished by detecting the order of fluorophores coupled to specific amino acids on a single TARGET peptide molecule (33). Such a technology has been developed by functionalizing a peptide with one type of fluorophore (Cy3) at the N-terminal site and a second type of fluorophore (Cy5) on an internal cysteine residue. The method then monitored the order in which the two fluorophores passed through Alexa488-labeled ClpP14 protease, as detected using the separation-dependent Forster resonance energy transfer effect (“FRET”). This approach provides less information than is potentially available by nanopore sequencing.

[0139] 3.9.4 Single molecule degradative sequencing.

[0140] Several methods have been developed for recognizing, detecting and removing one amino acid at a time from a peptide and either recording (e.g., in DNA) or detecting (e.g., by optical readout methods) the result, providing a single molecule version of classical Edman peptide sequencing. These methods have in common the need to recognize individual terminal amino acids, a problem that has not been completely solved to date, with the result that only an “approximate” version of a peptide’s sequence is obtained. As described below, even this approximate information may be sufficient to allow use of these detection technologies with the invention, since the invention requires discrimination among only a relatively small number of different TARGET peptides. The instrument platforms associated with these technologies typically provide for simultaneous “sequencing” of millions to billions of peptide molecules, with successive amino acids decoded in successive cycles of reagent recognition and terminal amino acid removal. Thus any of them can be used to generate sequence data (or “approximate” sequence data) from the peptide libraries prepared using the invention.

[0141] 3.9.4.1 Single molecule degradative sequencing by reverse translation.

[0142] The concept of “reverse-translation” of a peptide sequence into a DNA sequence is of course a noteworthy contradiction of the central dogma of molecular biology (DNA makes RNA makes protein), and no such biological system is believed to exist. However, by coupling the recognition of a specific terminal amino acid on an immobilized peptide (by a recognition molecule specific for one or more terminal amino acids) with a transfer of a DNA code from the recognition molecule to a nearby DNA molecule, one position of the peptide’s sequence can be converted into a DNA code. By then removing the terminal amino acid (e.g., by chemical Edman degradation, or limited enzymatic attack by an exoprotease, or other equivalent means) and repeating the recognition and transfer process, a DNA sequence can be progressively generated that encodes all or part of the peptide amino acid sequence (the peptide being destroyed in the process: i.e., the reading process is degradative). The DNA molecule can subsequently be read using any of the established DNA sequencing methodologies. In the event that some amino acids are not clearly discriminated, or that some recognition molecules only recognize a class of amino acids (e.g., those having positive charge, orthose with negative charge, or those with uncharged hydrophilic side chains, etc.), the resulting “approximate” sequence information may nevertheless be sufficient to recognize one peptide sequence among a limited set of expected alternatives. A variety of methods can be used to immobilize millions of individual peptide molecules and adjacent DNA molecules so as to produce a DNA library encoding sequence information from the original peptide library. In this technology, peptides are typically linked to the solid support via the c-terminal carboxyl group, leaving the n- terminus free. Significant progress towards reverse translation of peptide libraries has been reported by several groups, e.g., the “Proteocode” technology (https: / / www.encodia.com / technology: US 2021 / 0208150) and the “ProtSeq” technology ((34); patent publication US 2021 / 01022).

[0143] 3.9.4.2 Single molecule degradative sequencing using .fluorescence detection.

[0144] Other degradative single peptide molecule methods have been reported that make use of optical detection of fluorescent labels. In one such method, the dynamics of binding of recognition reagents to terminal amino acids of single peptide molecules located in individual wells on a semiconductor chip are sensed and interpreted peptide sequence data (35). In this technology, peptides are typically linked to the solid support via the c-terminal carboxyl group (PCT / US2021 / 028471).

[0145] Another method ("fluorosequencing”, (36-38)) makes use of fluorescent labels attached to specific amino acids (e.g., cysteine SH, lysine NH2, etc.) by chemical methods, and records the disappearance of these fluorescent signals when a labeled amino acid is cleaved off during a sequence of degradative (e.g., Edman) steps.

[0146] 3.10 SAMPLE BARCODING

[0147] The ability to manipulate DNA sequences, particularly to synthesize, sequence, and splice together DNA sequences of various lengths, enables the attachment of designed, recognizable sequence tags (i.e., “barcodes”) to the DNA recovered from biological samples. Once sample DNA is barcoded, then DNA from multiple samples can be combined for further processing, such as next generation sequencing (“NGS”), and afterwards attributed to the correct original sample (i.e, demultiplexed). A variety of DNA barcode systems have been developed with the object of reliable identification of the original source sample in NGS applications. Note: this use of the term barcode (meaning a designed tag used for labeling) is distinct from an alternative usage applied to endogenous DNA sequences found to be characteristic of biological species and used to identify presence of a species in a sample comprised of multiple organisms.

[0148] For high-throughput DNA sequencing applications, Xu, et al, devised a library of 240,000 orthogonal 25mer DNA barcodes in 2009 (39), and continuing work by a number of investigators has resulted in barcode libraries of improved reliability and error-resistance (40- 42). In these applications, where the analytical system has the ability to directly sequence the barcode along with the sample DNA, the barcodes can be relatively short, and sophisticated mathematical approaches for error-detection and correction can be employed to reduce the likelihood of incorrect barcode assignments due to incorrect base calls, insertions, deletions, etc.. Sophisticated software, including machine learning, has been developed to improve correct DNA barcode assignment in multiplexed nanopore DNA sequencing (e.g., (43)).

[0149] DNA barcodes are also used in other applications where the barcodes are “read” by hybridization of a complementary probe that can be detected by optical or other detection means (44) without sequencing (e.g., using a fluorescently labeled complementary-sequence probe reagent detected by single molecule microscopic imaging). Such methods have been successfully applied for single molecule fluorescence detection of up to 1,000 different mRNA sequences in single cell images using 16 different 30-mer readout probes in a 16-bit modified Hamming distance 4 code (45). Such coding methods enable efficient sample barcoding and demultiplexing in single molecule imaging platforms.

[0150] Potential errors in barcode identification resulting from sequencing errors, hybridization failures, etc., have been dealt with by application of techniques derived from information theory. A particularly effective approach has been the use of error-correcting codes (ECC) used in digital information storage systems (e.g., computed memory), generally stemming from the work of Hamming (46). By incorporation of extra parity bits in an encoded signal, Hamming codes can be designed to allow detection and repair of single-bit errors, and detection (and in some cases repair) of two bit errors. Given the potential for error in barcode readout by nucleic acid sequencing (e.g., in single molecule nanopore methods) or hybridization (e.g., in optical imaging applications), error detection and correction can be critical when single molecules are being detected and counted to determine a quantitative result.

[0151] Yet other barcoding strategies have been created using DNA barcodes to label synthetic chemical libraries (47) and peptide barcodes detected by mass spectrometry (48).

[0152] 3.11 MACHINE LEARNING FOR SIGNAL CLAS SIFIC ATION.

[0153] Machine learning methods have been successfully developed that allow the identities and / or sequences of individual molecules to be deduced from complex signal patterns. Nucleic acid sequences can be derived from current traces measured as DNA or RNA molecules pass through nanopores using highly-trained neural networks to recognize and interpret conductivity transitions (49). Proteins can be recognized by machine learning based on optically-detected stochastic binding of multiple promiscuous affinity reagents to single molecules (31). In general, machine learning approaches make it possible to improve the recognition of molecules by all the above single molecule technologies by building mathematical models based on large numbers of reference examples, and incorporating more data for each example than is practical in human-designed programs.

[0154] 3.12 LIMITATIONS OF EXISTING TECHNOLOGIES AND OBJECTIVES OF THE PRESENT INVENTION

[0155] The current dominant methods for direct detection of peptide molecules by MS, including SISCAPA and related methods, have significant limitations. These include A) sensitivity limited by the performance of available mass spectrometers (currently limited to 10-100 amol of peptide, equivalent to 6 million to 60 million molecules of a peptide); B) low throughput (largely due to the limited speed of typical liquid chromatography systems employed); C) lack of robustness of the liquid chromatography systems used to separate peptides and introduce them into the MS; D) level of expertise required to operate LC-MS systems); E) high cost of LC-MS systems and the consequent limited adoption in clinical laboratories and F) impracticality of use in low-technology environments. In addition, there is the fundamental limitation that MS typically resolves and identifies analytes based on one or a few parameters that are derived from the peptide sequence (typically its mass and the masses of one to three of its specific fragments), but it does not typically determine the entire peptide sequence and is therefore susceptible to various forms of identification error.

[0156] Existing methods of single molecule detection likewise have significant limitations restricting their application to peptide quantitation. These include A) absence of internal standards that could provide a quantitative reference for comparison of samples; B) limited dynamic range (inability to count sufficient numbers of molecules to estimate frequency of very low abundance targets); C) lack of efficient sample preparation protocols to deliver peptides for single molecule detection; and D) limited ability to recognize many of the amino acids in a peptide sequence (i.e., limited specificity).

[0157] Recognizing these limitations, a recent analysis (50) of the limitations of single molecule methods in comparison with mass spectrometry for the general characterization of complex proteomes concluded that the above-referenced limitations, and specifically the dynamic range limitation, effectively prevent current single molecule methods from providing achieving a general analysis of sample proteins.

[0158] It is an object of the present invention to transcend these limitations and others. The invention provides significant improvements in assay sensitivity by making use of single molecule counting technologies instead of mass spectrometry detection, with the potential to make quantitative measurements at the level of hundreds to thousands of analyte molecules (i.e., >1, 000-fold improvement compared to MS methods, including SISCAP A-MS). The invention provides sequence-based assay specificity through direct detection and counting of analyte molecules without the use of liquid-chromatography or expensive mass spectrometer instruments. The invention makes use of certain technologies and platforms that have been extensively developed for nucleic acid applications (e.g., DNA and RNA sequencing), some of which have been implemented commercially as small, inexpensive instruments capable of generating accurate results in low-technology environments. A further object of the invention is to significantly lower the cost of making precise measurements of protein biomarkers, drugs and targets, and thereby to enable expanded use of quantitative protein tests in diagnostics and in longitudinal health monitoring. The invention provides methods for improved protein quantitation by adapting a novel specific affinity enrichment strategy to allow detection of enriched peptides by technologies other than mass spectrometry - specifically technologies that enable counting individual peptide molecules in a sequence-specific manner. In adapting the specific affinity enrichment strategy to these alternative detection means, significant novel changes as described herein are required in the selection and treatment of peptides, in the generation of suitable internal standards as substitutes for stable isotope labeled versions, in the generation of sequence- specific binding reagents, in the preparation and delivery of peptides for single molecule detection, and in the analysis of resulting data.

[0159] The present invention will be described with respect to particular embodiments and with reference to certain drawings, but the invention is not limited thereto but only by the claims. It is to be understood that not necessarily all aspects or advantages may be achieved in accordance with any particular embodiment of the invention. Thus, for example those skilled in the art will recognize that the invention may be embodied or carried out in a manner that achieves or optimizes one advantage or group of advantages as taught herein without necessarily achieving other aspects or advantages as may be taught or suggested herein.

[0160] The invention, both as to organization and method of operation, together with features and advantages thereof, may best be understood by reference to the following detailed description illustrated in the accompanying drawings. The aspects and advantages of the invention will be apparent from and elucidated with reference to the embodiment s) described hereinafter. Reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment, but may. Similarly, it should be appreciated that in the description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various inventive aspects. This method of disclosure, however, is not to be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment.

[0161] It should be appreciated that "embodiments" of the disclosure can be specifically combined together unless the context indicates otherwise. The specific combinations of all disclosed embodiments (unless implied otherwise by the context) are further disclosed embodiments of the claimed invention.

[0162] DEFINITIONS OF TERMS

[0163] Key terms used frequently herein:

[0164] The term "amino acid" in the context of the present disclosure is used in its broadest sense and is meant to include organic compounds containing amine (NH2) and carboxyl (COOH) functional groups, along with a side chain (e.g., a R group) specific to each amino acid. In some embodiments, the amino acids refer to naturally occurring L amino acids or residues. The commonly used one and three letter abbreviations for naturally occurring amino acids are used herein: A=Ala; C=Cys; D=Asp; E=Glu; F=Phe; G=Gly; H=His; K=Lys; L=Leu; M=Met; N=Asn; P=Pro; Q=Gln; R=Arg; S=Ser; T=Thr; V=Val; W=Trp; and Y=Tyr (Lehninger, A. L., (1975) Biochemistry, 2d ed., pp. 71-92, Worth Publishers, New York). The general term "amino acid" further includes D-amino acids, retro-inverso amino acids as well as chemically modified amino acids such as amino acid analogues, naturally occurring amino acids that are not usually incorporated into proteins such as norleucine, and chemically synthesized compounds having properties known in the art to be characteristic of an amino acid. For example, analogues or mimetics of phenylalanine or proline, which allow the same conformational restriction of the peptide compounds as do natural Phe or Pro, are included within the definition of amino acid. Such analogues and mimetics are referred to herein as "functional equivalents" of the respective amino acid. Other examples of amino acids are listed by Roberts and Vellaccio, The Peptides: Analysis, Synthesis, Biology, Gross and Meiehofer, eds., Vol. 5 p. 341, Academic Press, Inc., N.Y. 1983, which is incorporated herein by reference

[0165] The term “analyte” may refer to any of a variety of different molecules, or components, pieces, fragments or sections of different molecules that one desires to measure or quantitate in a sample.

[0166] The term “anti-peptide antibody” (a class of specific binding agent, or BINDER) as used herein means a macromolecule capable of non-covalently and reversibly binding to a peptide in a manner that is specific to all or a portion of the peptide’s sequence. The term includes a variety of types of macromolecules as indicated in the definition of “antibody” above, and is not limited to the proteins conventionally considered antibodies.

[0167] The term “barcode” includes any distinguishing physical, chemical or sequence characteristic of a peptide construct capable of having multiple values that can be determined by a single molecule detection method. Nucleic acid sequences can be used as barcodes, for example by providing a set of distinguishable sequences, a different one of which can be linked to the peptides of each sample, identifying (“decoding”) the source of these peptides after the peptides from multiple samples are pooled for efficient processing in a single molecule detection system. Other forms of molecular barcodes can be used as well, including sets of glycan structures (which can be decoded using various specific lectins, for example), peptides (when these can be linked to and sequenced with the TARGET peptides); non-biological polymers distinguishable by length or content of alternative polymer units; and small molecules including colored or fluorescent dyes.

[0168] The term “bind” includes any physical attachment or close association, which may be permanent or temporary. Generally, reversible binding includes aspects of charge interactions, hydrogen bonding, hydrophobic forces, van der Waals forces, etc., that facilitate physical attachment between the molecule of interest and the analyte being measured. The “binding” interaction may be brief as in the situation where binding causes a chemical reaction to occur. Reactions resulting from contact between the binding agent and the analyte are also within the definition of binding for the purposes of the present invention, provided they can be later reversed. The terms “BINDER”, “antibody”, “anti-peptide affinity reagent”, “specific affinity reagent”, “specific binding reagent”, “affinity capture reagent” and “anti-peptide antibody” as used herein mean a reagent having the ability to reversibly bind to a specific TARGET peptide (and its cognate STANDARD) in a manner that is specific to all or a portion of the peptide’s sequence. Such a BINDER will typically bind a TARGET peptide with greater affinity, greater kinetic on-rate or lower kinetic off-rate than a majority of the other peptides present in samples, sample digests, or other sources of contamination. The terms include antibodies and fragments thereof as well as non-naturally occurring or synthetic antigen binding molecules. Thus, included are IgG antibodies (polyclonal, monoclonal, oligoclonal, etc.), and other antibody isotypes, fragments thereof, such as Fab fragments, murine, chimeric, and other non-human or not fully human antibodies and fragments thereof, synthetic (non-naturally occurring) antigen binding formats such as single chain antibodies and bispecific antibodies, as well as aptamers (including DNA, RNA and other polymeric aptamers) and binding proteins built from non- antibody structures (e.g., nanobodies).

[0169] The term “BINDER ID” means a molecular barcode identifying the BINDER to which a molecule bound in an enrichment step.

[0170] The term “biologic” means a drug produced by a biological mechanism, such as a protein; i.e., a protein therapeutic, or protein drug.

[0171] The term “biomolecules” refers to any molecule present in a biological system, and includes proteins, nucleic acids (specifically DNA and RNA in its various forms, both intracellular and extracellular), complex sugars (glycans and the like), lipids, and a variety of metabolites.

[0172] The term “denaturant” includes a range of chaotropic and other chemical agents that act to disrupt or loosen the 3-D structure of proteins without breaking covalent bonds, thereby rendering them more susceptible to proteolytic treatment. Examples include urea, guanidine hydrochloride, ammonium thiocyanate, trifluoroethanol and deoxycholate, as well as solvents such as acetonitrile, methanol and the like. The concept of denaturant includes non-material influences capable of causing perturbation to protein structures, such as heat, microwave irradiation, ultrasound, and pressure fluctuations. The term “click chemistry” means the use of pairs of chemical groups that react with each other but not with other chemical groups commonly found in biomolecules: i.e., they are bio-orthogonal coupling mechanisms. Commonly used click chemical pairs include, but are not limited to, a 3’ transcyclooctyne (TCO) group reacting bio-orthogonally with a tetrazine group (e.g., methyltetrazine (Me-TZ)), and bicyclononyne (BCN) reacting bio-orthogonally with an azide group. In some instances copper (Cu) ions serve as catalysts for a click reaction, and in other instances, typically involving a strained cyclic alkyne, a catalyst is not required.

[0173] The term “clonotypic” means uniquely characteristic of a clonal product, typically referring to a peptide sequence unique to a specific monoclonal antibody.

[0174] The term “cognate” as used herein means a relationship between molecules in which either 1) the molecules each contain a region that has the same structure as the other, or 2) the molecules can bind together by a specific interaction. In the case of peptides (e.g., a TARGET and STANDARD peptide pair), cognate peptides can share a region of identical sequence, which may be from 2 amino acids up to the full length. The difference between cognate peptides can be a difference in sequence, or a difference due to attachment or removal of some atom(s) or groups (including one or more entire amino acids), or the addition to the peptide or a chemical group of any size (including oligonucleotides, peptides, “handles” such as biotin, and reactive groups able to subsequently bond to other molecules).

[0175] The term “cognate BINDER” or “cognate affinity capture reagent” means a specific affinity reagent (e.g., a specific binding reagent, BINDER) that is capable of specifically binding a cognate TARGET peptide and / or cognate STANDARD, in the sense that the cognate affinity capture reagent is designed, generated or selected to have a specific affinity for an epitope comprising part or all of its cognate peptide sequence.

[0176] The term “degradative sequencing technology” means a technology in which peptide molecules are disassembled one amino acid at a time (or in some cases two amino acids at a time), typically from one end, and the terminal amino acid identified, e.g., as one of the 20 common amino acids found in proteins, or as one of a subset of amino acids. In some cases, the identification can be obtained directly by optical or electrical readout, and in some cases the amino acid identity is translated into another molecular form (e.g., DNA) for later readout using a different technology. The terms “drug” and “therapeutic” mean a type of molecule that may, under appropriate circumstances of dosing and timing, interact with components of a subject’s body to modify biological processes, including disease processes, normal processes, aging and the like. A drug may be a small molecule such as aspirin, or a macromolecule such as a protein (e.g., insulin), a nucleic acid (such as an anti-sense drug), or carbohydrate (such as heparin). Drugs that comprise or are derived from monoclonal antibodies represent a growing class of therapeutic agents with particular advantages in terms of extreme specificity for endogenous protein and other targets involved in disease processes.

[0177] The term “electrospray ionization” (ESI) refers to a method for the transfer of analyte molecules in solution into the gas and ultimately vacuum phase through use of a combination of liquid delivery to a pointed exit and high local electric field.

[0178] The term “elution” means the release of a bound peptide or construct from a BINDER.

[0179] The term “flag” is used herein as equivalent to Barcode, and may be any type of distinguishing molecular feature including, but not limited to, a polymer of dissimilar subunits encoding an identification relevant to sample analysis.

[0180] The terms “Forster resonance energy transfer” or FRET refer to energy transfer between two light-sensitive molecules (chromophores, typically fluorescent molecules). A donor chromophore, initially in its electronic excited state, may transfer energy to an acceptor chromophore through nonradiative dipole-dipole coupling. The efficiency of this energy transfer is inversely proportional to the sixth power of the distance between donor and acceptor, making FRET extremely sensitive to small changes in distance, generally on scales of 1 to 10 nm.

[0181] The term “immobilized enzyme” means any form of enzyme that is fixed to the matrix of a support by covalent or non-covalent interaction such that the majority of the enzyme remains attached to the support of the membrane.

[0182] The term “ligation” as used herein means the joining of an end of a polymer chain (such as a nucleic acid) to an end of another polymer chain to form a combined linear polymer. The term includes joining by enzymatic means (such as that of a DNA ligase, splicing means such as CRISPR, and other well-known molecular biology techniques for joining and splicing nucleic acid sequences) and chemical means (such as the use of click chemistry).

[0183] The term “Linkage” means a connection between originally separate molecules, and includes common covalent connections between units found in biopolymers and man-made polymers, as well as connections made using chemistries such as the well-known “click” chemistries, reactions such as those between amino groups and NHS esters, and formation of sugar-phosphate bonds when oligonucleotides are ligated together, as well as strong but non- covalent connections such as the interaction between biotin and streptavidin. The related term “Linker” means a segment of a molecule comprising an atomic configuration capable of, or arising from, formation of a linkage between two or more initially separate molecules.

[0184] The term “MALDI” means Matrix Assisted Laser Desorption Ionization and related techniques such as SELDI, and includes any technique that generates charged analyte ions from a solid analyte-containing material on a solid support under the influence of a laser or other means of imparting a short energy pulse.

[0185] The term “Mass spectrometer” (or “MS”) means an instrument capable of separating molecules on the basis of their mass m, or m / z where z is molecular charge, and then detecting them. In one embodiment, mass spectrometers detect molecules quantitatively. An MS may use one, two, or more stages of mass selection. In the case of multistage selection, some means of fragmenting the molecules is typically used between stages, so that later stages resolve fragments of molecules selected in earlier stages. Use of multiple stages typically affords improved overall specificity compared to a single stage device. Often, quantitation of molecules is performed in a triple-quadrupole mass spectrometer using the method referred to as ‘Multiple Reaction Targeting’ or “MRM mass spectrometry” in which measured molecules are selected first by their intact mass and secondly, after fragmentation, by the mass of a specific expected molecular fragment. However, it will be understood herein that a variety of different MS configurations may be used to analyze the molecules described. Possible configurations include, but are not limited to, MALDI instruments including MALDI-TOF, MALDI-TOF / TOF, and MALDLTQMS, and electrospray instruments including ESI-TQMS and ESLQTOF, in which TOF means time of flight, TQMS means triple quadrupole MS, and QTOF means quadrupole TOF. The terms “molecular tag”, “molecular flag”, or “molecular feature” mean a structural component of a molecular construct that can be detected by a single molecule detector and assigned a significance in the interpretation of counted molecules (e.g., distinction between TARGET and STANDARD tags, barcodes identifying samples, barcodes identifying BINDERS, etc.)

[0186] The terms “particle” or “bead” mean any kind of particle in the size range between lOnm and 1cm, and includes magnetic particles and beads.

[0187] The term “peptide library preparation” means a method used to convert the proteins in a biological sample into a collection of peptides modified so as to be detectable, identifiable and countable by a sequence sensitive single molecule detector.

[0188] The term “proteolytic treatment” or “proteolytic enzyme” may refer to any of a large number of different enzymes, including trypsin, chymotrypsin, LysC, ArgC, AspN, GluC, v8 and the like, as well as chemicals, such as cyanogen bromide, that, in the context of the methods described herein, acts to cleave peptide bonds in a protein or peptide in a sequence-specific manner, generating a collection of shorter peptides (a digest).

[0189] The term “proteotypic peptide” means a peptide whose sequence is unique to a specific protein in an organism, and therefore may be used as a stoichiometric surrogate for the protein, or at least for one or more forms of the protein in the case of a protein with splice variants.

[0190] The term “sample” means any complex biologically-generated sample derived from humans, other animals, plants or microorganisms, or any combinations of these sources. “Complex digest” means a proteolytic digest of any of these samples resulting from use of a proteolytic treatment.

[0191] The term “SAMPLE ID” means a molecular barcode identifying the sample from which a molecule was obtained, i.e., its sample of origin. A sample barcode present in a construct identifies the sample of origin and allows this identity to be recovered after constructs from multiple samples have been pooled and analyzed together in a single molecule detector.

[0192] The terms “ratchet mechanism”, “protein nanomachine” or “molecular motor” mean a molecular-scale device capable of pulling, pushing, unzipping (in the case of complementary strands of nucleic acids), or otherwise regulating the motion of linear molecules in discrete steps.

[0193] The term “sequence-sensitive single molecule detection” or “SSSMD” means detection and counting of individual molecules using a method capable of differentiating between different linear biopolymer sequences occurring in the molecules. A “sequence-sensitive single molecule peptide detector” means a detector, instrument, technology, chemistry, or multi-component system that is able to achieve sequence-sensitive single molecule detection of peptides. Such a detector need not achieve 100% accuracy to accomplish the objectives of the invention, since the number of different peptide sequences that must be distinguished from one another and counted in the invention is a small number (e.g., 1, 1-5, 1-10, 5-20, 10-50, 25- 100, 50-200, or more peptides) compared to number of peptides present in a digest of a complex biological sample (typically hundreds of thousands of peptides in the digest of a sample such as blood plasma). The term includes nanopore-based sequencing of nucleic acids, proteins and peptides; fluorescence-based methods such as fluorosequencing (36-38) including Edman methods; “reverse-translation” of peptide sequencing into DNA sequences followed by DNA sequencing (the “Proteocode” technology developed by Encodia: https: / / www.encodia.com / technology); “FRET” fingerprinting of peptides (36, 51) single molecule imaging methods (31) and other related methods.

[0194] The terms “sequencing nanopore” and “nanopore” as used herein refer to ion- conductive pores capable of functioning in an ion-impermeable membrane or vessel wall, and through which linear polymers can pass. Typical nanopores are of biological origin (e.g., MspA), comprising one or more protein molecules, or created by engineering (e.g., versions of biological nanopores modified by mutation, rearrangement or combination of proteins; very small holes etched or drilled in thin metallic or ceramic substrates; or DNA assemblies). A recording of the current flowing through a nanopore over time is referred to as a “trace” or “squiggle”.

[0195] The term “sequential degradation” refers to a process in which amino acid residues are removed, in sequence order, from one terminus of a peptide. In the context of the invention, sequential degradation can be employed in a process in which a peptide’s terminal amino acid is “recognized” (e.g., by binding of one of a series of affinity regents specific for the various amino acids presented at the terminus) and its identity determined or recorded for later evaluation, after which the terminal amino acid can be cleaved off (e.g., using enzymes such as exoproteases, classical Edman chemistry, or other chemistries capable of removing a terminal amino acid) and the process repeated to determine a sequence of amino acids from the peptides’ terminus. Similarly, a process can employ recognition reagents that report information on two or more terminal amino acids at a time, and a cleavage process can be employed that removes two or more terminal amino acids per cycle. The process need not sequence all amino acids in a peptide to generate TARGET peptide or STANDARD identifications and single molecule counts that are useful in the invention.

[0196] The term “SISCAP A” means the method described in US Patent No. 7,632,686, and in Mass Spectrometric Quantitation of Peptides and Proteins Using Stable Isotope Standards and Capture by Anti-Peptide Antibodies (SISCAP A) (Journal of Proteome Research 3: 235-44 (2004).)

[0197] The term “small molecule” or “metabolite” means a multi-atom molecule other than proteins, peptides and DNA; the term can include but is not limited to amino acids, steroid and other small hormones, metabolic intermediate compounds, drugs, drug metabolites, toxicants and their metabolites, and fragments of larger biomolecules.

[0198] The term “stable isotope” means an isotope of an element naturally occurring or capable of substitution in proteins or peptides that is stable (does not decay by radioactive mechanisms) over a period of a day or more. The primary examples of interest in the context of the methods described herein are C, N, H, and O, of which the most commonly used are 13C and 15N.

[0199] The term “solubilized tissue sample” means a liquid sample generated from a sample of a solid biological tissue (e.g., liver, brain, skin, etc.) by a method that results in a solution containing tissue molecules. Depending on the method of solubilization, a solubilized tissue sample may contain one, a few, many or almost all tissue molecules in solution. Tissue solubilization can be achieved by a variety of methods including grinding, pulverization, ultrasonication, homogenization, and similar mechanical methods, as well as exposure to liquid solutions including detergents, solvents, protease inhibitors, salts, buffers, and the like. The terms “STANDARD”, “internal standard peptide”, “internal standard”, “labeled TARGET”, or “labeled TARGET peptide” may be any altered version of the respective TARGET fragment or TARGET peptide that is 1) bound by the appropriate BINDER with an affinity and kinetics very similar to that with which the cognate TARGET fragment or TARGET peptide is bound, and 2) differs from it in a manner that can be distinguished from the cognate TARGET peptide by a sequence-sensitive single molecule peptide detector (e.g., by means of some sequence difference, amino acid modification, inclusion of a non-natural chemical group), or a mass spectrometer (either through direct measurement of molecular mass or through mass measurement of fragments, e.g., through MS / MS analysis), or by another equivalent means. In the case of a nanopore detector, for example, a suitable TARGET peptide and its STANDARD would produce distinguishable ion current signatures while passing through the nanopore.

[0200] The term “STANDARD tag” or “STANDARD flag” means a molecular tag or feature within or attached to a STANDARD peptide enabling a single molecule detector to distinguish the STANDARD tag from a TARGET tag. The STANDARD tag may be the absence of any TARGET tag. A STANDARD tag may consist of the absence of a feature present in a cognate TARGET tag, the presence in the STANDARD tag of a feature absent in the cognate TARGET tag, or the presence of different features in the STANDARD tag and TARGET tag. Multiple different STANDARD tags may be used, provided that the STANDARD tags are distinguishable from any TARGET tags.

[0201] The term “standardized sample digest” or “standardized sample” means a protein or peptide sample to which one or more STANDARD version(s) of one or more TARGET peptide or protein analytes have been added in an amount that is a) known (in terms of concentration, mass, moles or other physical units) or b) consistent between samples (allowing quantitative comparison of TARGET peptide amounts between samples even if the absolute amount of the STANDARD added is not known). Once the sample is standardized with respect to a given TARGET peptide, then the ratio between TARGET peptide and STANDARD represents and preserves information concerning the amount of the TARGET peptide in the sample, allowing this information to be recovered by later quantitative analysis even if a variable amount of the TARGET peptide and STANDARD pair is recovered during a suitable enrichment process (i.e., a process that does not distinguish between the TARGET peptide and STANDARD peptides) prior to analysis.

[0202] The term “stoichiometric” refers to relationships between quantities of different molecules. In some chemical contexts, the word stoichiometry refers to presence of different elements or compounds in simple integral ratios, as prescribed by an equation or formula. Thus a TARGET peptide sequence that occurs once in the sequence of a parent protein target has a 1 : 1 stoichiometric relationship with the target, and can therefore be used as a quantitative surrogate to measure amounts of the protein. In the more general sense used in this disclosure, stoichiometry means a ratio relationship between molecules (or elements) that may have any numerical value, including non-integer values. In biological samples, two different proteins in blood or in a cell can have a relative stoichiometry extending over a very broad range, in principle from one molecule (the lower limit) of a low abundance protein to hundreds of billions of molecules (or more) for a high abundance protein in the same sample.

[0203] The terms “stoichiometric flattening”, “normalization”, “equalization” or “differential enrichment” refer to processes by which different molecules (e.g., peptides) that are present in a sample (e.g., a biological sample digest) at different concentrations (or in different amounts in mass or molar terms) are brought closer to equal concentrations or amounts. An example of such a process is an affinity enrichment method in which a larger relative fraction of a low abundance molecule is captured while a smaller relative fraction of a higher abundance molecule is captured (e.g., by adjusting the amounts of the corresponding affinity reagents, such as antibodies, used to accomplish this capture), the captured molecules being then separated from the sample and released from capture, resulting in a more nearly equivalent amount of the molecules in the processed sample. In order to preserve information as to the relative abundances of the molecules in the original sample, internal standard versions of the molecules (e.g., STANDARD versions of TARGET peptides) are added before this enrichment step, and both TARGET peptide and STANDARD measured in the resulting enriched sample. Given an estimate of the amount of STANDARD added to the sample, the ratio of TARGET peptide to STANDARD can be used to calculate the TARGET peptide abundance in the original sample. The terms “Subject” or “Patient” means a biological individual such as an individual human being or an animal.

[0204] The terms “TARGET” or “TARGET peptide” means a peptide chosen as a TARGET fragment of a protein or peptide. The TARGET may be any piece of a protein or peptide which can be produced by a reproducible fragmentation process (e.g., digestion using a proteolytic enzyme, or without a fragmentation if the TARGET fragment is the whole analyte) and whose abundance or concentration can be used as a surrogate for the abundance or concentration of the analyte.

[0205] The term “TARGET tag” or “TARGET flag” means a molecular tag or feature within or attached to a TARGET peptide enabling a single molecule detector to distinguish the TARGET tag from a STANDARD tag. The TARGET tag may be the absence of any STANDARD tag. A TARGET tag may consist of the absence of a feature present in a cognate STANDARD tag, the presence in the TARGET tag of a feature absent in the cognate STANDARD tag, or the presence of different features in the TARGET tag and STANDARD tag. Multiple different TARGET tags may be used, provided that the TARGET tags are distinguishable from any STANDARD tags.

[0206] The term “tag” is used herein as equivalent to Barcode, and may be any type of distinguishing molecular feature including, but not limited to, a polymer of dissimilar subunits encoding an identification relevant to sample analysis.

[0207] The term “T / S tag” or “T / S flag” means either a TARGET or a STANDARD tag, or a set of mixed TARGET and STANDARD tags, as will generally be present in a standardized sample.

[0208] The term “VEHICLE” means a molecule (for example a polymer such as an oligonucleotide or a polyethylene glycol, a linker comprising chemically reactive sites such as NHS or click chemistry groups, or a macromolecular carrier such as a bead or a “SNAP” particle), to which TARGET and STANDARD peptides (together with their associated distinguishing tags) can be linked in order to facilitate single molecule detection. A VEHICLE can include barcodes identifying a sample of origin, barcodes identifying a BINDER used to enrich specific cognate TARGET and STANDARD peptides. VEHICLES can also include one or more additional molecular structures that facilitate the transport of TARGET and STANDARD peptides to a single molecule detector, their presentation to such a detector, their transport through a detector (such as a nanopore), or their immobilization to a site or in a region observed by such a detector.

[0209] The use of the singular herein in any instance (e.g., “a” construct or “a” peptide), unless otherwise indicated, is intended to mean one or more and is not intended be limited to only one.

[0210] 5 THE PRESENT INVENTION

[0211] The inventions herein provide improved quantitative measurement of proteins, and peptides derived from them, through improvements to previous methods including the replacement of mass spectrometric detection by other detection techniques capable of identifying and counting individual molecules.

[0212] In the descriptions that follow, quantitation of proteins, peptides and other biomolecules is addressed in a general sense, and hence the invention disclosed is in no way limited to the analysis of blood, plasma and other body fluids.

[0213] 5.1 BRIEF DESCRIPTION OF THE DRAWINGS :

[0214] Figure 1 : Examples of abundances, TARGET peptide and STANDARDS for proteins in human plasma

[0215] Figure 2: Adding Amino Acids to a TARGET to Create a STANDARD

[0216] Figure 3: Rope-Tow Constructs with STANDARD and TARGET Sequence Tags

[0217] Figure 4: Two-Step Rope-Tow Constructs with Double Tags

[0218] Figure 5: TARGET / STANDARD Barcoding Followed by Enrichment

[0219] Figure 6: Two-step Digestion for Differential Modification of Peptide Ends

[0220] Figure 7: Rope-Tow Constructs with Sequencing Adapters: Ligation of rope-tow constructs with nanopore sequencing adapters

[0221] Figure 8: Multi -Epitope Binders Figure 9: DNA Sample Barcodes (16-bit example)

[0222] Figure 10: Decoding Peptide-A and Sample Barcodes

[0223] Figure 11 : Identifying Peptides Using Epitopes

[0224] Figure 12: Assembly of a Peptide:Oligo Construct for Nanopore Detection

[0225] Figure 13: Details of a Peptide:Oligo Construct for Nanopore Detection

[0226] Figure 14: Scheme for double ligation of tryptic peptides with c-terminal lysine (i.e., two amino groups) using “Click” chemistry

[0227] Figure 15: Click Ligation of Tryptic (Lys) Peptides to Motor Assemblies

[0228] Figure 16: Peptide Loop Insertion with Enzymatic Cut: Insertion of TARGET Peptide Into Oligo VEHICLE as a Loop, Followed by Sequence-Specific Enzymatic Oligo Cleavage

[0229] Figure 17: Peptide Loop Insertion with Chemical or UV Cleavage

[0230] Figure 18: Peptide Loop Preparation of TARGET and STANDARD

[0231] Figure 19: Parallel TARGET and STANDARD Constructs

[0232] Figure 20: Specific Affinity Capture of TARGET and STANDARD

[0233] Figure 21 : Peptide Loop: Effect of Failure to Insert Fully

[0234] Figure 22: Nanopore sequencing of constructs

[0235] Figure 23 : Rope-tow Constructs

[0236] Figure 24: Concatenation of constructs by hybridization and ligation

[0237] Figure 25: Constructs Prepared Using Bi-functional Supports

[0238] Figure 26: Example Rope-Tow Construct: Detailed example structure of a rope-tow peptide: oligo construct

[0239] Figure 27: Concatenation of Tryptic (Lys) Peptides Using Click Chemistry

[0240] Figure 28: Concatenated constructs

[0241] Figure 29: Rope-Tow Ligation with Splints

[0242] Figure 30: Ligation of Double-Tag Rope-Tow Constructs Figure 31 : Analysis of Detection Events in Affinity Imaging Detection

[0243] Figure 32: Identifying Peptides: Before and After Cleavage

[0244] Figure 33: Stoichiometric Flattening

[0245] Figure 34: Equalization: Stoichiometric Flattening

[0246] Figure 35: Multiplex Test Panel for SARS-CoV-2: A multiplex combination of molecules detectable using nanopore sequencing, including a) peptides from the SARS-CoV- 2 NCAP protein; b) SARS-CoV-2 Spike and NCAP protein linear epitopes whose binding by patient antibodies indicates vaccination or exposure to the virus; c) proteotypic peptides of three proteins used as plasma biomarkers of inflammation; and d) the RNA genome of SARS-CoV-2.

[0247] Figure 36: Design and Nanopore Analysis of Loop-Insertion Constructs

[0248] Figure 37: Nanopore Traces of Loop-Insertion Constructs

[0249] Figure 38: Preparation of Constructs for Reverse Translation Detection

[0250] Figure 39: Assembly of a Peptide:Oligo Construct for Reverse Translation Detection

[0251] 6 DETAILED DESCRIPTION OF THE INVENTION:

[0252] 6.1 SUMMARY OF THE INVENTION.

[0253] 1. A molecular construct and vehicle comprising:

[0254] (a) a molecular construct comprising a peptide comprising a target peptide sequence derived from proteolytic cleavage of a target protein and a molecular tag defining the source of said peptide, and

[0255] (b) a vehicle capable of presenting the construct for analysis by a sequence-sensitive single molecule detector.

[0256] 2. The molecular construct and vehicle of paragraph 1, wherein the molecular tag is a target tag that identifies the peptide as a peptide created by proteolytic digestion of a biological sample. The molecular construct and vehicle of paragraph 1, wherein the peptide comprises a synthetic peptide and the molecular tag is a standard tag that identifies the synthetic peptide as an internal standard. The molecular construct and vehicle of paragraph 1, wherein the vehicle is capable of binding said construct either to a support or to a soluble adapter capable of presenting the construct for analysis by a sequence-sensitive single molecule detector. The molecular construct and vehicle of paragraph 2, wherein more than 90% percent of the target molecules present in said sample digest are linked to target tags. The molecular construct and vehicle of paragraph 2, further comprising a SAMPLE barcode identifying the sample of origin. The molecular construct and vehicle of paragraph 1 further comprising a BINDER barcode identifying a binder to which the construct has been bound. The molecular construct and vehicle of any of the preceding paragraphs wherein the barcode or the tag is an oligonucleotide. The molecular construct and vehicle of any of the preceding paragraphs wherein the sequence-sensitive single molecule detector comprises a nanopore, a single molecule imaging system, or a single molecule degradative peptide sequencer. A plurality of reagents, comprising: the molecular construct and vehicle of paragraph 3, a tag reagent capable of reacting with a target peptide in a proteolytic digest of a biological sample to create the molecular construct of paragraph 2, and a binder that binds to said molecular constructs of paragraph 2 and paragraph 3 with similar affinity and kinetics. The plurality of reagents of paragraph 10, wherein the binder contacting a standardized sample digest comprising the molecular construct and vehicle of paragraph 2 and paragraph 3 binds the molecular construct of paragraph 2 and paragraph 3 in a ratio equal within 2%, 5%, 10% or 20% to the ratio in which they are present in said standardized sample digest. The plurality of reagents of paragraph 10, further comprising: one or more reagents capable of proteolytic fragmentation of sample proteins, and one or more solid supports for binders, including magnetic beads, non-magnetic beads, porous supports usable in packed columns, or chemical reagents capable of introducing reactive groups into peptides. A calibrator sample for peptide quantitation by a sequence-sensitive single molecule detector, comprising an amount of the molecular construct and vehicle of paragraph 2 and paragraph 3 in a known ratio. The calibrator sample of paragraph 13, wherein at least one of the constructs is present in known amount or concentration. A standardized sample digest derived from a proteolytic digest of a biological sample, comprising: an amount of a molecular construct comprising a target tag and a target peptide, said construct being a target peptide construct and an amount of a molecular construct comprising a standard tag and a peptide whose sequence is the same or similar to the sequence of said target peptide, said construct being a standard peptide construct, wherein the target peptide is generated by proteolytic digestion of a target protein in said biological sample, wherein said target and standard tags can be distinguished by a single molecule detector and comprise chemical or structural groups covalently joined to peptides in their respective constructs, wherein said target tag is covalently attached to a plurality of the peptides present in said sample digest, wherein said target peptide construct comprises more than 90% of the target peptide molecules present in said sample digest and wherein said standard peptide construct is prepared separately and added to said digest in a known amount, or in a consistent relative amount across a multiplicity of samples. The standardized sample digest of paragraph 7, wherein the number of molecules of the standard peptide construct added to the sample digest differs by no more than a factor of 100 from the number of molecules of the target peptide construct in said sample digest. The standardized sample digest of any one of paragraphs 15-16, further comprising one or more additional standard peptide constructs having a different standard tag from each other and with each construct at a different relative abundance. The standardized sample digest of any one of paragraphs 15-17, wherein the target tag is covalently attached to a majority of the peptides generated by proteolytic digestion of said sample. The standardized sample digest of any one of paragraphs 15-18, wherein said tags are oligonucleotides, An enriched standardized sample digest, comprising a bound fraction of the standardized sample digest of any one of paragraphs 15-16 bound by a binder, wherein said bound fraction comprises a target peptide construct and a standard peptide construct in a ratio equal within 2%, 5%, 10% or 20% to the ratio in which they are present in said standardized sample digest. A stoichiometrically-flattened standardized sample, comprising a plurality of pairs of cognate standard and target peptide constructs enriched from a standardized proteolytic digest of a biological sample by binding to their respective cognate binders, wherein a pre-enrichment ratio calculated by dividing the number of molecules of a first target peptide construct that is the most numerous of said target peptide constructs in the standardized sample digest by the number of molecules of a second target peptide construct that is the least numerous of said target peptide constructs in the standardized sample digest is more than 10 times larger than a post-enrichment ratio calculated by dividing the number of molecules of said first target peptide construct by the number of molecules of said second target peptide construct in said enriched sample. A method for the measuring the amount of a selected target protein in a biological sample, comprising: proteolytically digesting said sample, modifying a plurality of peptides in the digested sample by adding a target tag to form a plurality of constructs comprising a selected target peptide derived from, and proteotypic of, said target protein, said plurality of constructs being target construct molecules, adding an amount that is known and / or consistent between a set of samples of a prepared standard peptide construct that is a cognate of said selected target peptide construct and comprises a standard tag, forming a standardized digest, enriching said cognate target and standard peptide constructs by contacting said standardized digest with a cognate binder, forming bound constructs, separating said bound constructs from unbound constructs to form enriched constructs, releasing said enriched constructs from said binder, linking said enriched constructs to a vehicle capable of presenting said enriched constructs to a sequence-sensitive single molecule detector, counting said enriched target construct molecules and said enriched standard construct molecules using a sequence-sensitive single molecule detector capable of distinguishing said target and standard tags and identifying said peptides, calculating the amount of said protein in said sample. The method of claim 22, wherein the calculating is performed by multiplying the amount of standard construct added by the ratio of the number of target construct molecules counted to the number of standard construct molecules counted by said detector. the method of claim 22 or 23, wherein, independently or in any combination:

[0257] • said binder is attached to a support,

[0258] • said linking of said vehicle to said constructs occurs while said constructs are bound to said binder,

[0259] • said presenting of enriched constructs to said detector occurs while said constructs are bound to said binder,

[0260] • peptides are bound to the binder, the binder washed, and the peptides eluted in two or more successive cycles of enrichment,

[0261] • said proteolytic digestion comprises at least two sequential steps resulting in peptide cleavage at different sites, and wherein peptides are covalently modified between two such steps (or wherein said first sequential step cleaves at lysine residues),

[0262] • said tags are added to said peptides by reaction with peptide amino groups,

[0263] • said tags are added to said peptides by chemical reaction at a single site in an n- terminal amino acid,

[0264] • said tags are added to said peptides by chemical reaction at a c-terminal lysine residue,

[0265] • said constructs comprise a non-peptidic component attached to an n-terminal amino acid and a different non-peptidic component attached to a c-terminal lysine residue,

[0266] • said proteolytic digestion comprises at least two sequential steps resulting in peptide cleavage at different sites, and wherein peptides retain an unmodified n- terminal amino group when presented to said detector,

[0267] • a sample barcode is linked to said constructs encoding the identity, or relative position within a sample set, of said standardized samples; a plurality of said standardized samples is pooled; said sample barcodes associated with construct molecules are read using a sequence-sensitive single molecule detector; and the counts of target and standard construct molecules for each sample are separated based on said sample ID barcode identifying the sample from which they were enriched, and wherein said barcode may be an oligonucleotide,

[0268] • a binder barcode is linked to said constructs identifying the binder by which they were enriched, and wherein said barcode may be an oligonucleotide,

[0269] • said construct molecules are joined together into concatamers prior to presentation to said detector,

[0270] • said detector determines all or part of the amino acid sequence of the peptide components of said construct molecules by a stepwise degradative process (or wherein the sequence of said target peptide is encoded in a nucleic acid component linked to target tags, standard tags, sample tags, binder tags and vehicles, and is read, and counted, by conventional DNA sequencing),

[0271] • said detector recognizes and decodes the target and standard constructs comprising target peptides, tags and optional additional barcodes using time- dependent variations in an electrical or optical parameter measured while the construct molecules transit a nanopore (or wherein said nanopore is biological, e.g., the common protein nanopores occurring in nature or derivatives thereof, or nucleic acid constructs, or a hole in a solid state inorganic material, e.g., Si3N4, SiO2, graphene, or MoS2) (or wherein said target and standard constructs comprise non-peptide polymers including nucleic acids that engage with a molecular motor, e.g., a polymerase or helicase to regulate the speed at which the constructs move through a nanopore (or wherein the nanopore detection is continued, and construct counts accumulated, until reaching pre-determined threshold numbers of counts, which may be based on counts required for each peptide sequence, e.g., to provide a pre-determined precision according to counting statistics, or counts required to achieve a pre-determined precision in a ratio, e.g., target / standard counts),

[0272] • said constructs are located on a support and detected by sequential binding of a plurality of binders comprising a detectable label, wherein independently or in any combination: o peptides in constructs are identified by using cognate binders labeled with optically detectable moieties including fluorescent dyes or proteins o multiple binders recognize distinct epitopes within a target peptide o kinetic binding analysis of binder-construct interactions is used to improve the specificity of detection o detection of binding or lack of binding by binders is interspersed with sequential removal of n-terminal amino acids or peptide segments o one or more of the said target tags, standard tags, sample tags, binder tags, or vehicles is an oligonucleotide and is detected by hybridization of an optically-labeled complementary oligonucleotide. A method for the measuring amounts of a plurality of selected target proteins in a biological sample, comprising independently or in any combination: proteolytically digesting proteins in said sample to yield peptides, modifying said peptides by covalent chemical addition of a target tag to form a plurality of constructs, including target construct molecules comprising selected proteotypic target peptides derived from said target proteins, adding prepared standard constructs that are cognates of said target constructs and comprise a standard tag in amounts that are known and / or consistent between a set of samples, forming a standardized digest, enriching said cognate target and standard construct pairs by contacting said standardized digest with cognate binders, forming bound constructs, separating said bound constructs from unbound constructs to form enriched constructs, releasing said enriched constructs from said binder, linking said enriched constructs to a vehicle capable of presenting said enriched constructs to a sequence-sensitive single molecule detector, recognizing and counting said enriched target construct molecules and said enriched standard construct molecules using a sequence-sensitive single molecule detector capable of distinguishing said target and standard tags and identifying peptides, calculating the amount of said proteins in said sample by multiplying the amount of each standard construct added by the ratio of the number of cognate target construct molecules counted to the number of cognate standard construct molecules counted by said detector, wherein a pre-enrichment ratio calculated by dividing the number of molecules of a first target peptide construct that is the most numerous of said target peptide constructs in the standardized sample digest by the number of molecules of a second target peptide construct that is the least numerous of said target peptide constructs in the standardized sample digest is more than 10 times larger than a post-enrichment ratio calculated by dividing the number of molecules of said first target peptide construct by the number of molecules of said second target peptide construct in said enriched sample.

[0273] The inventions herein provide improved quantitative measurements of the amounts of proteins, in a way that is highly specific, extremely sensitive, multiplexable with wide dynamic range, capable of very high throughput with low cost per measurement, and amenable to implementation on compact, inexpensive equipment.

[0274] The invention combines a series of known and new processes in a novel combination, and provides novel advantages over existing protein measurement methods. In its most basic aspect, the invention comprises proteolytic fragmentation of target protein(s), addition of internal standard versions of one or more peptides in known amount (these standard peptides being detectably different from the sample digest peptides based on incorporation of molecular tags into either sample digest peptides, added standard peptides, or both), enrichment of selected sample peptides (TARGETs) and cognate internal standards (STANDARDS) by specific affinity selection on BINDERS, and single molecule identification and counting of the resulting enriched peptides. By counting individual peptide molecules, the invention provides the maximum sensitivity attainable by direct analyte detection (i.e., detection without amplification). The following general features, and others described herein, are included in the invention:

[0275] Proteolytic digestion of proteins in a sample to yield a peptide digest.

[0276] Selection of TARGET peptides from among the candidate peptides produced by digestion of a target protein, based on theoretical (e.g., in silico) and / or experimentally determined features and performance as a quantitative surrogate of the protein analyte.

[0277] Design of internal standard (STANDARD) versions of TARGET peptides (which may be identical sequences).

[0278] Coupling TARGET molecular tags to TARGET peptides to form TARGET constructs, coupling STANDARD molecular tags to STANDARD peptides to form STANDARD constructs, or both.

[0279] Addition of known or reproducible amounts of STANDARD constructs to a digest prior to enrichment, creating a standardized sample digest. Specific enrichment of TARGET and STANDARD peptide constructs, removal from digest matrix, washing, chemical modification as needed, and elution, creating an enriched standardized sample digest. This specific enrichment step may be carried out with amounts of BINDERS adjusted to achieve some degree of stoichiometric flattening among a series of TARGET peptides, creating an enriched and flattened standardized sample digest.

[0280] Optional addition of BINDER-identifying barcodes to TARGET and STANDARD peptide constructs in a sample digest.

[0281] Optional addition of sample-identifying barcodes to TARGET and STANDARD peptide constructs in a sample digest when samples are to be pooled prior to single molecule detection. Sample barcode addition can be carried out either before or after specific enrichment by BINDERS.

[0282] Presentation of enriched TARGET and STANDARD peptide constructs to a sequence- sensitive single molecule detector in an appropriate chemical form or structure, typically by linkage to a VEHICLE.

[0283] Detection and counting of peptide TARGET and STANDARD molecules using the sequence- sensitive single molecule detector.

[0284] Estimation of the amount(s) of TARGET peptide(s) (and thus parent proteins) in a sample based on the ratios between respective cognate TARGET and STANDARD peptide counts, with pooled samples decoded when necessary using sample barcodes.

[0285] Several technologies among those known in the art can be used for sequence-specific single molecular detection and counting of biological macromolecules. Some of these have been developed for peptide sequencing, while others have been developed primarily for sequencing nucleic acids, but show potential for application to peptides.

[0286] The present disclosure provides methods for preparation and analysis of protein- containing biological samples compatible with any of the sequence-sensitive single molecule detection methods.

[0287] In some embodiments the invention provides a means to measure the amount of a peptide molecule (termed a "TARGET” peptide), typically a proteolytic fragment of a sample protein resulting from proteolytic digestion of a biological sample. In some embodiments, sample proteins are proteolytically digested and standardized by addition of an internal standard peptide or peptide construct (STANDARD) to create a standardized sample, from which the TARGET peptide and STANDARD are enriched and individually counted using sequence-sensitive single molecule detection (e.g., nanopore sequencing). The resulting counts of TARGET peptide and STANDARD molecules allow estimation of the absolute or relative amount or concentration of the target protein in the sample, given knowledge of the amount of STANDARD present in the digest.

[0288] 6.2 COMPONENTS USED IN THE INVENTION

[0289] Four molecular components used in some embodiments of the invention are described below: a) one or more TARGET peptides (the analytes to be measured); b) internal standard (STANDARD) versions of TARGET peptides or peptide constructs used as internal standards; c) specific affinity reagents (BINDERS) for capture and enrichment of TARGET peptides and STANDARD prior to detection; and d) barcodes used to distinguish TARGET and STANDARD constructs and / or to distinguish peptides on constructs derived from different samples. In general, these components are selected, prepared or optimized in ways surprisingly distinct from earlier work in which mass spectrometry has been used for peptide detection.

[0290] 6.3 TARGET PEPTIDE.

[0291] Using the known sequence of a target protein, one or more peptide segments within it are selected as “TARGET” peptides. Selection can be accomplished using an “in silico” approach (e.g., by “digesting” the sequence of a target protein, known for example from the genome sequence of the relevant species, using a computer to cut the sequence at sites predicted based on the known cleavage specificities of a selected protease or chemical fragmentation method), an experimental approach (e.g., from a list of peptide fragments actually observed in a digest of the protein or a sample containing it), or both. A preferred TARGET peptide can be defined in some embodiments by criteria selected from a set intended to identify peptides that optimize the performance of the assay. In some embodiments it is preferred that a TARGET peptide has one or more characteristics that improve its performance in an assay according to the invention, including, but not limited to, peptides that are or have:

[0292] A length of about 4 to 50 amino acids, or more preferably 6 to 24 amino acids; Efficient digestion: Produced rapidly and in high (ideally >90%) yield by digestion of the target protein through an efficient, inexpensive proteolytic treatment with an enzyme (e.g., a protease such as trypsin) or a high-yield chemical treatment (e.g., CNBr cleavage);

[0293] Proteotypic sequence: A sequence that is unique to the target protein (unless measurement of a family of proteins is an object of the assay), i.e., that it is "proteotypic" for the protein and appears in no other proteins likely to be found in the intended sample (or more preferably, that it occurs in no other natural protein coded for by the genome of the species of interest), and that it occurs in the target protein sequence in a known number of locations (typically one location, but potentially more than one if the peptide sequence is repeated in the protein);

[0294] Few variants: A sequence that is relatively consistent across the population of target protein molecules occurring in a sample, i.e., occurs with relatively few sequence variants (unless measurement of such sequence variants is an object of the assay).

[0295] Few post-translational modifications: Peptides with relatively few post-translational modifications (unless measurement of such modifications is an object of the assay)

[0296] Favorable epitopes: A sequence containing structural features (e.g., “immunogenic” epitopes in the case of antibody affinity reagents) that facilitate development of specific affinity reagents capable of binding the peptide with high affinity, and specifically a slow off- rate (e.g., monoclonal antibodies, aptamers, etc.).

[0297] Solubility: Favorable physico-chemical properties, including solubility in aqueous solutions, little or no binding to materials used in sample preparation and analysis vessels and devices (e.g., nanopores), and little or no tendency to aggregate.

[0298] Stability: Low rate of spontaneous chemical degradation (e.g., by methionine or tryptophan oxidation, asparagine or glutamine de-amidation, etc.).

[0299] Recognizable sequence: A sequence that has features making it easily distinguishable from other sequences using the chosen sequence-sensitive single molecule detection means. If, for example, the detection means is nanopore sequencing, then peptide sequences that produce distinctive current traces as molecules pass through a nanopore, thus allowing them to be distinguished from other peptides, are preferred. Typical nanopores produce current signals that are reflective of a stretch of 3-6 contiguous amino acids (a “kmer”) inside the pore. Amino acids have a multiplicity of different side-chain volumes, and while these volumes do not always directly determine the nanopore “blockade currents”, sequences with more variable patterns of side chain volume are preferred. If the detection means involves recognition and recording of a terminal amino acid and its removal in a cyclical process to expose a new terminus (e.g., using degradative sequencing technology), then sequences that include amino acids for which the terminal recognition is most accurate, and / or least confusing, are likely to be preferred. If the detection means involves binding of recognition molecules to peptide epitopes, then peptides with multiple distinct epitopes are preferred.

[0300] Amino acid constraints: Presence or absence of specific amino acids that impact assay performance. In some embodiments cysteine is avoided due to its potential to form bridges between peptides, or alternatively a step is included in the sample preparation to block cysteines (e.g., via alkylation by iodoacetamide). In some embodiments, one or more cysteine residues present in a peptide are used as reactive sites for introduction of linkages to other molecules, or labels that can assist in peptide recognition by a sequence-sensitive single molecule detector.

[0301] Amino groups: Specific numbers and sites of amino groups (lysine side chain and n-terminal). As the most preferred sites for chemical linkage to other molecules, the number and position of amino groups is an important factor in the design of some constructs required for efficient presentation of peptides to a sequence-sensitive single molecule detector. For example, a tryptic peptide with c-terminal arginine and no internal lysine residues has a single amino group located at its n-terminus (i.e., it is “single amino”), and therefore has a unique site at which certain amino-reactive chemistries can establish a covalent linkage between the peptide and another molecule. Alternatively, a tryptic peptide having a c-terminal lysine and no internal lysines has two amino groups: one at the n-terminus and one at the epsilon amino group of the lysine side chain, thus facilitating methods in which a peptide is linked into a longer polymer by coupling at both ends. Carboxyl groups: Specific numbers and sites of carboxyl groups (glutamic and aspartic acid side chains and at the c-terminus). As an alternative preferred site for chemical linkage to other molecules, the number and position of carboxyl groups is an important factor in the design of some constructs required for efficient presentation of peptides to a sequence-sensitive single molecule detector.

[0302] Electric charges: Specific numbers of charged amino acids (e.g., lysine, arginine, histidine, and the n-terminus with positive charges; and glutamic and aspartic acids and the c- terminus with negative charges) and the sum of these (i.e., the net charge of the peptide at the working pH of a sequence-sensitive single molecule detector). The total charge of a peptide can significantly affect its movement through a nanopore under the influence of an electric potential between cis and trans compartments, between which the nanopore serves as a conduit: in some embodiments a net negative peptide charge is preferred, such that the peptide is pulled through the pore from cis to trans (i.e., in the same direction a negatively charged oligonucleotide would be). In some embodiments a net positive peptide charge is preferred, such that a peptide is dragged through the pore by another molecule to which it is attached (and which has a net negative charge). In some such embodiments, it is preferred that positive charge(s) be localized towards the end of the peptide that is last to enter the pore (i.e., the trailing end of the peptide) so as to help maintain the peptide in an extended, linearized form as it passes into and through a nanopore.

[0303] These characteristics vary considerably, and in many cases independently, across the population of potentially selectable TARGET peptides to which the invention may be applied, as well as between single molecule detection technologies, and as a result the selection of optimal TARGET peptide sequences appropriate for some embodiments involves a complex weighting of these and other characteristics.

[0304] 6.3.1 Distinction from selection for mass spectrometry

[0305] It is useful to note that in many embodiments, commonly-considered features of proteolytic peptides in conventional peptide quantitation methods are NOT included among peptide characteristics considered important or limiting in the present invention: these include i) performance of a peptide in chromatographic separations (e.g., favorable elution behavior in reversed-phase liquid chromatography, elution time separated from other peptides used in panel combinations, etc.); ii) ionization efficiency in a mass spectrometer source (including the preference for a c-terminal positive charge); iii) fragmentation behavior in a mass spectrometer (important for example when using triple quadrupole “MRM” quantitation methods), and iv) peptide size (i.e., as a limitation due to the preferred mass range of a mass spectrometer). In view of the inapplicability of these criteria, some embodiments of the present invention will make use of TARGET peptides different from those preferred in mass spectrometry-based detection systems (including peptides that would be unusable in MS detection).

[0306] 6.3.2 Peptides with 1 or 2 amino groups

[0307] In some embodiments TARGET peptides are generated through cleavage of proteins by trypsin (an inexpensive and well-understood protease that cuts polypeptide chains preferentially c-terminal to lysine and arginine residues). Tryptic TARGET peptides can be selected to contain either 1 (“Single amino”) or 2 (“Double amino”) amino groups. Numerous alternative proteases (e.g., Lys-C, Arg-C, pepsin, papain, chymotrypsin, etc.) and chemical cleavage reactions (cyanogen bromide (CNBr) cleaving at methionine (Met) residues; BNPS- skatole cleaving at tryptophan (Trp) residues; formic acid cleaving at aspartic acid-proline (Asp-Pro) peptide bonds; hydroxylamine cleaving at asparagine-glycine (Asn-Gly) peptide bonds, and 2-nitro-5-thiocyanobenzoic acid (NTCB) cleaving at cysteine (Cys) residues) can also generate peptides with characteristics allowing their use as TARGET peptides.

[0308] In some embodiments, peptides with a single amino group are preferred because this provides a unique and chemically convenient site that can be covalently coupled to other molecules used in enrichment or detection of peptides. Tryptic peptides ending in c-terminal arginine (R) and having no internal lysine(s) possess only one amino group (at the n-terminus) and are termed “Single amino” peptides. Reaction with this group provides a geometrically well-defined “handle” on one end of the peptide. Such peptides are used in some embodiments to create “rope-tow” constructs for nanopore sequencing through combination with oligonucleotides, as described below. In some embodiments, it is preferred that the TARGET peptide has either no net charge or a net positive charge in order to facilitate peptide movement within a nanopore. Peptides of the “Single amino” group may be selected to contain no aspartic or glutamic acids so as to minimize the contribution of negative charges on the peptide (i.e., the peptide preferably has zero or net positive charge that, in some embodiments, helps resist being pulled through a nanopore by its attachment to a negatively charged polymer construct), and no cysteine (so as to avoid the necessity for a method step to block these reactive groups). Single amino peptides with no aspartic or glutamic acid residues also have a single carboxyl group, which is useful in some embodiments that rely on anchoring a peptide to a support via its c-terminus, leaving the unmodified n-terminus available for binding of affinity reagents or sequential degradation (e.g., by Edman chemistry).

[0309] In some embodiments requiring linkage sites on both ends of a peptide, it can be convenient to select TARGET peptides (e.g., “Double amino” peptides) with a single lysine residue present at or near the c-terminus, whose epsilon-amino group provides a second reactive amino group, in addition to the n-terminal amino group at the opposite end of the peptide molecule. Linkage through these two amino groups allows a peptide to be coupled “in-line” with preceding and succeeding polymers via amine-reactive chemistry to form a continuous thread. Peptides with an internal lysine, or multiple lysines, are less preferred in such embodiments due to the potential for multiple non-linear constructs.

[0310] In some embodiments TARGET peptides are selected for other configurations of reactive sites. In some embodiments it is preferred that a peptide have a single carboxyl group and this criterion can be met by peptides with no aspartic or glutamic acid residues while possessing a free c-terminal carboxyl. In some embodiments other specific amino acids are desired so as to facilitate labeling of peptide molecules with amino acid-specific detection reagents.

[0311] In some embodiments in which a net positive peptide charge is preferred, aspartic or glutamic acid carboxyl groups in the peptide can be converted to positive charged sites by a chemical modification, e.g., activation by a carbodiimide and reaction with a reagent having an amino group (that couples to the carboxyl) and a second positively charged group.

[0312] In some embodiments, peptides having post-translational modifications are selected. For example, the n-terminal amino group of tryptic peptide VHLTEEPK from the beta chain of human hemoglobin (Hb) is modified by glycation in a fraction of molecules in the blood as a result of slow reaction with blood glucose (the modified Hb is referred to as HbAlc, and is used clinically as a measure of average blood glucose over time in a test for diabetes). The unmodified form of this peptide is thus “Double amino”, while the modified form is “Single amino”. In some embodiments designed to measure the fraction of Hb that is modified by this glycation, the n-terminal amino group of unmodified peptide is first blocked by reaction under conditions favoring reaction with the lower pK n-terminal amino group, and subsequently both modified and unmodified forms of the peptide are coupled to other molecules by the single remaining amino group of the c-terminal lysine.

[0313] 6.4 STANDARD VERSION OF A TARGET PEPTIDE FUNCTIONING AS AN INTERNAL STANDARD.

[0314] 6.4.1 History in mass spectrometry

[0315] In mass spectrometry (MS) as used for peptide quantitation (52), an internal standard is typically a synthetic, same-sequence version of a TARGET peptide including one or more amino acids comprising stable isotope labels (typically referred to as a Stable Isotope Standard or SIS) that allow it to be distinguished from the sample-derived TARGET peptide by mass measurement in the MS instrument (i.e., the well-known method of “isotope dilution”). Given the chemically identical structures of TARGET peptides and stable isotope labeled peptide internal standards, there is no basis on which to suspect that specific capture reagents (e.g., anti-peptide antibodies or other BINDERS) can distinguish between them, and thus such capture reagents will bind TARGET and stable isotope labeled peptides in whatever ratio they exist in the surrounding solution (e.g., the sample digest). A key feature of this approach to internal standardization is that the same method can be used to create the standard in all cases: for example, with tryptic peptides an effective standard can be made by synthesizing the TARGET sequence with a c-terminal amino acid (typically either lysine or arginine) containing stable isotopes (e.g., all 12C replaced by 13C and all 14N replaced by 15N). Therefore no advanced design or experimental testing and selection is required in a particular case: one approach works in all cases. This is not the case in the present invention, in which some embodiments require an involved selection and / or manufacturing strategy to identify and produce cognate TARGETS, STANDARDS and BINDERS that function properly together in the invention. 6.4.2 Limited use of internal standards in nucleic acid-based technologies

[0316] Internal standards that are truly identical, or almost identical (i.e., “cognate”), to the target analyte sequence are not typically used in genomics and nucleic acid technologies. Since the genome has very flat stoichiometry to begin with (most genes present in equal copy numbers, or if not, present in small integer ratios), internal standardization is typically not required for copy number quantitation. In quantitative methods such as RNAseq, where quantitation is important and mRNA levels can vary widely, current approaches typically sequence deeply and use total counts of specific sequences without differential enrichment, while focusing on between- sample normalization methods to reduce the effect of systematic factors that influence the results, such as RNA length, GC content, structure formation potential, etc. Where nucleic acid quantitation is important (e.g., in detection of one or more specific sequences such as SARS-CoV-2 sequences), PCR and related technologies are typically employed that rely on amplification, and the result is expressed in terms of the number of amplification cycles required to achieved a certain detection threshold (e.g., a “Ct value”). In none of these cases is a true internal standard (one chemically equivalent to the analyte in the assay but distinguishable by the detector) required or employed.

[0317] Since most of the emerging single molecule peptide technologies are evolutions of genomic methods, and / or are aimed at biomarker discovery rather than precise quantitation over a wide dynamic range, internal standardization with cognate STANDARDS has not previously been a significant objective of single molecule peptide analysis by these methods. It is thus a novel object of the present invention to provide effective internal standardization for single molecule peptide methods so as to enable accurate peptide quantitation by single molecule methods.

[0318] 6.4.3 Need for novel internal standards for quantitative single molecule detection

[0319] In the present invention, a single molecule detection technology is not expected to be able to reliably detect small mass differences (a few atomic mass units) between the otherwise identical chemical structures of a TARGET peptide and an isotopically-labeled STANDARD with the same sequence (i.e., same chemical structure), and therefore other differences in molecular character besides isotopic mass must be employed. In this case a difference in chemical structure is required, and it would be expected based on experience with the extreme specificity of high-affinity specific capture reagents that a structural difference would in general lead to a significant difference in binding by a specific capture reagent, which in turn would interfere with the ability of a chemically modified peptide to function as an effective internal standard by preserving the ratio of TARGET to STANDARD during enrichment by BINDER from the digest.

[0320] In some embodiments of the invention, a STANDARD is identified and prepared for each TARGET peptide and added to the sample in known or constant amount before, during or after digestion, but before enrichment, to act as a quantitative reference at the detection step. A sample digest to which STANDARDS corresponding to cognate TARGETS have been added in known or constant amount is referred to herein as a “standardized sample digest”. A sample digest may be standardized with respect to a single TARGET, or with respect to multiple TARGETS. The amount of a TARGET peptide can be compared with the amount of added STANDARD, and thereby measured, by multiplying the amount of STANDARD by the observed ratio of TARGET peptide to STANDARD in a sample. In some embodiments, the STANDARD is very similar to the TARGET peptide, i.e., as close as possible to being indistinguishable from it during steps of the workflow before the detection step, while being clearly distinguished from it at the detection step - in other words a cognate sequence peptide standard herein referred to as a STANDARD.

[0321] In some embodiments, the STANDARD serves as an internal standard against which the TARGET peptide amount is compared, for example by comparing the number of TARGET peptide molecules to the number of STANDARD molecules, providing a ratio measurement. In some embodiments in which a known amount of STANDARD is added (e.g., a known mass, or molar amount, or known number of molecules), multiplication of the ratio by this amount yields the amount of TARGET peptide (or mass or the number of TARGET peptide molecules) in the sample digest. In the case where the same amount (although not necessarily an amount whose mass in moles, grams or molecules, or concentration is known) of STANDARD is added to multiple samples, the presence in each of a consistent amount of STANDARD allows the amounts of TARGET peptide to be compared between these samples (using the shared amount of STANDARD as basis) to provide relative quantitation within a sample set. Samples can be compared using this approach by addition of the same amount of STANDARD (i.e., the same mass or the same volume of the same or equivalent solution), or using different amounts in different samples so long as the amounts added to different samples are known in relative terms (e.g., twice as much STANDARD added to sample 2 and to sample 1). Methods for the use of STANDARDS at various levels of monitor peptide quantitation and calibration are described in detail in Provisional patent application 63 / 213,371 - Calibration of Analytical Results in Dried Blood Samples, which is incorporated by reference herein in its entirety.

[0322] It is advantageous for a STANDARD construct and the STANDARD peptide it comprises to be as similar as possible to the respective TARGET construct and the TARGET peptide it comprises, since this similarity minimizes the probability that the ratio between them (which encodes the desired quantitative result of the analytical process) will be skewed or altered by some physical or chemical process in any step of an analytical workflow prior to detection, including enrichment by a cognate BINDER. As noted above, since a BINDER selected to enrich cognate TARGET and STANDARD constructs (or the respective peptides) must be highly specific in order to bind these peptides and not the enormous variety of other peptides present in a digest of a biological sample, some embodiments of the invention make use of TARGET and STANDARD peptides that are identical (i.e., perfect cognates). Alternative approaches, in which limited modifications of peptide sequence or structure distinguish TARGET and STANDARD peptide components, are less ideal and less general, but in some cases may be practically useful.

[0323] Similarly, for TARGET and STANDARD constructs, where a structural or chemical distinction is required in order that they be separately countable by a single molecule detector, the non-peptidic components of the TARGET and STANDARD constructs should also be cognates, though with relaxed similarity constraints. It is therefore advantageous for the non- peptidic components of the TARGET and STANDARD constructs to have similar physical properties such as mass, physical dimensions, shape, charge, hydrophobicity, solubility, etc. In some embodiments of the invention these constraints are addressed by the use of oligonucleotide TARGET and STANDARD tags, wherein the tags have the same length and may have the same base composition (implying the same molecular mass), but different sequences, allowing them to be distinguished by DNA sequencing or by specific hybridization to complementary probes. Linkage of such oligo tags to TARGET to STANDARD peptides can be accomplished using bifunctional linkers (for example including flexible polymer components such as polyethylene glycol between the oligo and peptide attachment sites) that reduce any steric hindrance the oligo may exert on the peptide that could affect binding to a BINDER. Such a level of similarity reduces the probability of skewing of the TARGET to STANDARD construct ratio because of differences in the diffusion, charge repulsion, epitope- masking, or solubility of the two constructs.

[0324] Suitably similar TARGET and STANDARD constructs form a cognate construct pair. The TARGET and STANDARD constructs, together with a BINDER capable of binding them without skewing the ratio, form a set of cognate molecules. A STANDARD construct, the cognate BINDER, a TARGET tag and any linker required to link the TARGET tag to TARGET peptide molecules in a sample digest form a cognate reagent set useful for specific measurement of the TARGET peptide and its parent target protein in a sample (i.e., they can serve as a kit for measuring the protein).

[0325] Achieving the goal of construct cognate equivalence nevertheless remains challenging because of the interdependence of constraints governing cognate constructs and BINDERS, and the absence of successful attempts to solve this problem in the past.

[0326] 6.4.4 Standards with altered TARGET amino acid sequence

[0327] In some embodiments, the STANDARD is created by replacement or alteration (e.g., by chemical modification) of one or more amino acids in the TARGET peptide sequence, or by addition of amino acids or other chemical structures. In some embodiments it is preferred that the replacement, addition or alteration a) does not result in any significantly difference in binding of the TARGET peptide and the cognate STANDARD to the cognate BINDER, and b) results in an easily detected change in the result from a sequence-sensitive single molecule detector (e.g., a different ion trace during transit of a nanopore compared to the TARGET peptide, or a different amino acid sequence detected by a degradative sequencing process, or a difference in the set of epitope-specific binders detected by a single molecule imaging platform).

[0328] In some embodiments, one or more amino acids or other chemical groups can be added to either the n-terminal or c-terminal end of the TARGET peptide to create a STANDARD, with the same constraints (e.g., an easily detected change in the result of a sequence-sensitive single molecule detector, but no significantly difference in binding of the TARGET peptide and STANDARD to the cognate specific BINDER). In some embodiments, these replacements and / or modifications are made to residues outside the peptide epitope to which a selected BINDER binds - such epitopes are typically linear contiguous regions of 4-8 amino acids in the case of IgG antibody BINDERS, leaving numerous potential modification sites available outside this region in a TARGET peptide 8-25 amino acids long.

[0329] In some embodiments a single serine (S) residue may be added to either the n-terminus or c-terminus of the sequence of a TARGET peptide to create a cognate STANDARD. Any other amino acid, or sequence of amino acids, that is clearly recognized by a sequence-sensitive detector can in theory be used instead of serine, the choice of added amino acid(s) being free, constrained only by the requirements of STANDARDS generally (i.e., BINDER binding equivalent to that of the cognate TARGET peptide, etc.) and any sequence constraints arising from any chemistry required to present peptides for detection. In some embodiments that make use of coupling reactions at two linkage sites on the peptide provided by the n-terminal amino group and the epsilon amino groups of a c-terminal lysine residue, addition of a residue after the lysine (a serine residue in this embodiment) provides a STANDARD that is chemically identical to the TARGET peptide along the entire chain of connected atoms between the peptide’s two amino groups (the n-terminal amino and lysine epsilon amino group) while comprising an appended serine residue “side chain”. In some embodiments, an amino acid such as serine can be added to the n-terminus. In other embodiments any amino acid(s) or chemically linkable group of atoms can be added to one or the other terminus, to an internal amino acid, or to both termini, to create a STANDARD version of a TARGET peptide sequence.

[0330] Figure 2 illustrates the challenge in practice of designing a simple addition of amino acids to the n-term or c-term of a peptide TARGET to create a cognate STANDARD while preserving equivalent binding to a BINDER. Here a series of versions of the peptide LLGPHVEGLK (proteotypic for human mesothelin) was synthesized with each of the 20 amino acids added to the n-terminus, and dipeptides added to the c-terminus (in each case a proline was added after the lysine and ahead of the variable amino acid in order to prevent removal of the added amino acid by trypsin cleavage, were the STANDARD added to a sample prior to digestion as is often the case). Each variant was mixed with a similar amount of the unmodified TARGET (LLGPHVEGLK), and the ratio of variant candidate STANDARD and TARGET signals measured by mass spectrometry before and after enrichment by a rabbit monoclonal antibody with specific affinity for this peptide. All n-terminal additions result in a dramatic decrease in binding of candidate STANDARDS compared to the TARGET: none are enriched to more than 4% of the level of the TARGET. The epitope recognized by this antibody thus probably includes the n-terminus and cannot accommodate an added amino acid. C-terminal additions are successfully enriched, with recoveries compared to TARGET of 27% (-PC) to 5386% (-PW); i.e., widely varying depending on the specific amino acid added after the proline. The antibody binding therefore appears to be affected by c-terminal additions, and in some cases (e.g., -PW) these c-terminal variants bind in preference to the original TARGET against which the antibody was made. Only 2 of the 39 variants examined bind the TARGET and candidate STANDARD at near-equivalence: c-terminal -PP and -PQ additions bind with approximately 99% and 102% recovery relative to the TARGET sequence. This example demonstrates that a large majority of modifications made by adding amino acids to the n-term or c-term of a 10 amino acid long TARGET peptide are unlikely to yield STANDARDS that bind equivalently to a given BINDER, preserving the TARGET / BINDER ratio present in the standardized sample digest. In addition, those STANDARD versions that appear to satisfy this simple version of the equivalence requirement (-PP and -PQ in this case), must pass further tests of equivalence under varying solution conditions, workflow timelines, and sample matrices, further restricting the range of choices.

[0331] In some embodiments using nanopore sequencing, and in which peptides transit the pore from n-terminal to c-term, it can be advantageous to use a STANDARD with an added residue at the c-terminus (the end closest to a DNA motor on the cis-side of the pore) so as ensure that the STANDARD variation is read, even if the peptide is longer than the nanopore’ s read depth and as a result some of the peptide’s n-terminal residues are not read during a period of controlled movement of the peptide through the nanopore. Sets of STANDARDS generated by addition of a constant c-terminal residue or residue pair to form the STANDARDS will in general require accurately reading a minimal subsequence of 2 amino acids more than the minimum required to distinguish the TARGETS themselves. Given the likelihood of imperfect reads, and the potential contamination with other, un-selected peptides, longer reads of 5, 6, 7, 8, or 10 amino acids, or the entirety of the peptide’s sequence may be required to identify the TARGET peptide and STANDARD molecules with sufficient accuracy (e.g., 99.5%, 99%, 98%, or 95% accuracy) to enable use of the ratio TARGET-to- STANDARD molecule counts to calculate a precise estimate of TARGET peptide amount.

[0332] In some embodiments, for example degradative methods in which a peptide is immobilized by the c-terminus and read by successively removing amino acids from the n- terminus, an added residue indicating STANDARD status is preferred at the n-terminus so as to ensure that the distinction between TARGET and STANDARD peptides is read at the beginning, and does not require sequencing to the end of the entire peptide. Accurately reading a minimal subsequence of 3 amino acids starting from the n-terminus is often sufficient to distinguish among a small set (e.g., 20) of TARGET peptides and their respective STANDARDS. Given the likelihood of imperfect reads, and the potential contamination with other, un-selected peptides, longer reads may be required to confidently identify the TARGET peptide and STANDARD molecules. In some embodiments, for example those that employ a sequential enzymatic (e.g., exoprotease) or chemical (e.g., Edman) process to remove single amino acid residues from one terminus of a peptide, the advantage of rapid definitive identification of TARGET and STANDARD sequences based on just a few terminal residues is substantial, since it could allow early termination of the cyclical read process, thus leading to a significant decrease in the number of cycles required and thus in analysis time, with associated decreases in cost and increased throughput.

[0333] In some embodiments, larger numbers of TARGET peptides and STANDARDS are used and need to be discriminated: for example, 25, or 50, or 100, or 200, or 400, or 600, or 800, or 1,000 TARGET peptides and their cognate STANDARDS; in such cases, based on an analysis of the uniqueness of the sequences, it may be desirable or required that a detector determine more of the peptide sequence, up to a complete sequence of some or all of the peptides.

[0334] In some embodiments, initial studies can be undertaken in which the TARGET peptides are sequenced beyond 3 or 4 residues, up to complete sequences, in order to detect the presence of any interfering peptides (i.e., peptides that share short n-terminal or c-terminal sequences with TARGET or STANDARD sequences, or otherwise generate output that can be confused with the pre-selected TARGET and STANDARD sequences) likely to be present in a given sample type. If such interfering sequences are commonly detected, deeper sequencing can be applied to distinguish the interfering sequences from TARGET or STANDARD sequences (i.e., sequencing up to or beyond the amino acid residue where the interfering peptide is no longer identical to a TARGET or STANDARD sequence), an approach with a high probability of success given that TARGET peptides are typically selected to be proteotypic in the species of interest.

[0335] STANDARDS generated by modification of a TARGET amino acid sequence face several challenges that motivate exploration of alternative approaches. These include rarity (the low probability of finding a modified sequence that binds equivalently to a cognate BINDER); lack of generality (the fact that each TARGET and cognate BINDER represent a separate case that must be individually optimized); and the fact that only some single molecule detection technologies are likely to be able to detect such a sequence difference reliably.

[0336] 6.4.5 Chemically modified peptides as standards

[0337] In some embodiments one or more amino acid residues of a TARGET peptide may be modified to generate a STANDARD. A large number of non-canonical amino acids that are known in the biochemical literature can be substituted for residues of the TARGET peptide or added to its sequence. Likewise, a large number of naturally occurring chemical modifications of amino acids are known and can be introduced into residues of the TARGET peptide during or after synthesis to form a STANDARD. Likewise, a large number of artificial chemical modifications can be made to amino acids of the TARGET peptide to form a STANDARD. Two examples of small but significant modifications are terminal blockages: 1) acetylation of an n-terminal amino group or 2) amidation of a c-terminal carboxyl group, both of which can be carried out easily during synthesis of a STANDARD peptide having the same sequence as a cognate TARGET, and both of which represent small alterations in the peptide structure. These small alterations can be “read” at a later stage of a single molecule workflow by reaction of peptides with a chemical reagent capable of efficiently combining with exposed amino or carboxyl groups (respectively). In some embodiments making use of amino groups to link peptides to oligonucleotides, blockage of a STANDARD’S c-term carboxyl can prevent reaction with a reagent that nevertheless reacts with a TARGET’S c-terminus: if the result of reaction with the reagent (which may for example add polymer or other structures to the TARGET’S structure) is detectable by a single molecule detector such as a nanopore, then the distinction between TARGET and STANDARD required by the invention can be provided. Any of these modifications may be used to create a STANDARD, provided that it meets the criteria described above (equivalent binding to a specific enrichment reagent, and equivalent reactivity in any required chemical reactions involved in sample preparation).

[0338] In some embodiments, for example those that employ nanopore sequencing, TARGET and STANDARD molecules may be “read” completely during passage through a nanopore, reducing the potential for confusion between expected TARGET and STANDARD sequences, or with potentially interfering sequences. In some nanopore sequencing embodiments capable of halting the reading of a peptide after reading a small number of amino acids and ejecting the peptide from the nanopore based on confidently identifying it as a specific TARGET or STANDARD sequence, the uniqueness of n-terminal or c-terminal sequences remains important and provides an opportunity to reduce time spent on unproductive sequence reading and therefore increase throughput of molecule counting.

[0339] 6.4.6 Discrimination of STANDARDS and TARGETS using members of different “click” chemical pairs

[0340] In some embodiments, STANDARDS and cognate TARGETS share an identical amino acid sequence but differ in an attached chemical group. For example, STANDARDS can comprise a peptide sequence linked to one member of a pair of “click” chemistry groups (e.g., TCO, capable of reacting bio-orthogonally with molecules comprising a tetrazine group, the other member of the click pair, or vice versa), while cognate TARGETS comprise the same (or very similar) peptide sequence linked to one member of a different pair of “click” chemistry groups (e.g., BCN, capable of reacting bio-orthogonally with molecules comprising an azide group, the other member of that click pair, or vice versa). Because the components of the two click pairs generally react only with the other pair member, but not between click pairs, such click-activated TARGETS and STANDARDS are generally inert until they encounter a molecule comprising the opposite pair member, at which time they spontaneously react forming a covalent linkage. Such click-activated TARGETS and STANDARDS are therefore each capable of reacting specifically with different additional molecules (e.g., oligonucleotides comprising the appropriate different click groups) at a later stage of a sample preparation workflow. In some embodiments, TARGETS and STANDARDS comprise different chemical linkage groups (e.g., selected from the above-mentioned click pairs) connected to the peptide by similar or identical spacers (e.g., polyethylene glycol of length 1, 2, 3, 4, 5 or more polymer units) thus reducing any potential impact of the difference in chemical linker structures (e.g., TCO and BCN as mentioned above) on the relative binding of TARGETS and STANDARDS to a cognate BINDER.

[0341] 6.4.7 Identification of STANDARDS by linkage to non-peptide flags or barcodes

[0342] In some embodiments a peptide is attached to another molecule to label it as a TARGET vs a STANDARD, to barcode it (e.g., to identify the sample from which it came), to facilitate or regulate its passage through a nanopore, or a variety of other purposes useful in a single molecule detection workflow. In some embodiments the distinction between TARGETS and STANDARDS is encoded in an attached, non-peptidic “tag” component rather than in the peptides’ structures themselves or in chemical linkage groups (e.g., click groups) they comprise. In some embodiments this is accomplished by preparing the STANDARD prior to its addition to a sample digest in a form that is already attached to a detectable tag (e.g., a nucleic acid sequence tag) that specifically indicates its status as a STANDARD. In an example of such an embodiment shown in Figure 3 A, an oligonucleotide VEHICLE comprises a 5’ phosphate 52 (to facilitate ligation with other nucleic acid chains), a preceding sequence 29, a residue 33 (indicated by X) capable of forming a linkage 34 with a terminal residue of peptide 52 (in this case a STANDARD peptide having the same sequence as a cognate TARGET), an abasic stretch 36 running alongside the peptide (forming a rope-tow construct as described herein), and a following sequence 30 comprising a tag sequence 54 (indicated by a box) that identifies the construct as containing a STANDARD peptide. In the example shown, the peptide GFVEPDHYVVVGAQR is a member of the class of “single amino” peptides, and thus comprises only a single amino group which is located at its n-terminus. Cognate TARGET peptide 53 (example shown in Figure 3B) in such an embodiment is attached to a VEHICLE of similar overall structure as the STANDARD construct, but comprising a different nucleic acid sequence tag 55 that indicates its status as TARGET. During single molecule detection, the VEHICLE nucleic acid sequences can be read by a nanopore and their location in relation to a peptide (e.g., preceding or following with pre- determined proximity) can be used to identify each peptide molecule as a TARGET or STANDARD molecule. The overall similarity of the VEHICLES attached to the pre-prepared STANDARD (Figure 3A) and sample digest-derived TARGET peptides (Figure 3B) minimizes any potential difference in binding of the peptide portions of the constructs (TARGET and STANDARD peptides 52 and 53 being structurally identical) to cognate peptide sequence-specific BINDERS. In some embodiments the sequence tags distinguishing the TARGET and STANDARD VEHICLES are optimized for high sequence accuracy in a given sequence-specific detection system (e.g., a nanopore reading system, or an affinity reagent imaging system).

[0343] The primary function of STANDARD tags and TARGET tags is to distinguish peptide constructs added to a sample as internal standards (STANDARD constructs) from peptide constructs that incorporate peptides created by proteolytic digestion of the sample proteins (TARGET constructs). In some embodiments the TARGET tag may be the absence of any STANDARD tag. In some embodiments the STANDARD tag may be the absence of any TARGET tag. A STANDARD tag may consist of the absence of a feature present in a cognate TARGET tag, the presence in the STANDARD tag of a feature absent in the cognate TARGET tag, or the presence of different features in the STANDARD tag and TARGET tag. In some embodiments, such presence / absence features may include differences in the sequence of oligonucleotide tags. The importance of maintaining unbiased (unskewed) ratio relationships between TARGET and STANDARD constructs (i.e., preserving their cognate character, specifically in regard to interaction with a cognate BINDER) argues against large structural differences between the TARGET and STANDARD tags (e.g., presence vs absence of a sizable chemical group). Multiple different STANDARD tags may be used, provided that the STANDARD tags are distinguishable from any TARGET tags. Multiple different TARGET tags may be used, provided that the TARGET tags are distinguishable from any STANDARD tags.

[0344] In some embodiments each different peptide STANDARD is prepared attached to a respective cognate VEHICLE that comprises a nucleic acid sequence tag that specifically identifies that STANDARD amino acid sequence and distinguishes it from a plurality of other STANDARDS that may be used in the same workflow. Use of different nucleic acid sequence tags for each STANDARD provides an orthogonal method for identifying these peptides, and this information makes it possible to assess the reliability of both methods (i.e., the degree of agreement between the peptide detection and tag detection), and to optimize the respective detection methods to improve accuracy.

[0345] In some embodiments the oligo tag sequences used to identify and distinguish cognate STANDARDS and TARGETS in a cognate group are selected so as to be chemically very similar (e.g., same length and base composition) while being reliably distinguishable (e.g., different base sequence). By being chemically similar, and not located in close proximity to the peptide, the tags are unlikely to have any differential effect on the binding of cognate TARGET and STANDARD molecules to the cognate BINDER, thus preserving the ratio of TARGET to STANDARD in the standardized digest. By having distinguishable sequences, as detected by any of the single molecule methods herein, the TARGET and cognate STANDARD molecules can be identified and counted reliably, thus providing an accurate value for the ratio of TARGET to STANDARD.

[0346] In some embodiments STANDARD-VEHICLE constructs (e.g., Figure 3A) are prepared and added to the sample digest after sample digest peptides have been incorporated into similarly-structured TARGET constructs (e.g., Figure 3B). In these embodiments, the structure of the STANDARD peptide molecule may be identical to the TARGET peptide structure (e.g., it can be a synthetic version of a known cognate TARGET peptide sequence), while their respective STANDARD and TARGET VEHICLES comprise distinct nucleic acid sequence tags (54 and 55 in Figure 3), thus ensuring that the cognate BINDER will bind the attached peptides equivalently, and thereby accurately preserve the TARGET -to- STANDARD ratio present in a standardized sample digest.

[0347] Given such an encoding scheme to identify STANDARDS and TARGETS the method requires that these peptides be joined to their respective tags prior to enrichment and selection using the cognate BINDERS. A specific advantage of this approach is the identical structures of cognate TARGET and STANDARD peptides, and thus the high likelihood that the cognate BINDER binds them with identical affinity and kinetics, thus preserving the TARGET-to- STANDARD ratio present in the sample digest. The potential disadvantage of this approach is the expense involved in using sufficient TARGET vehicles to incorporate all the digest peptides. A further simplification of this approach to address this issue is described below.

[0348] Some embodiments make use of a further simplification of the VEHICLE-encoding scheme shown in Figure 3 and described above. Figure 4 shows a method in which a short identifying oligo a STANDARD tag 62 is attached to a STANDARD peptide 52, in this case by an amine-reactive N-hydroxysuccinamide (NHS) group 61 attached by linker 34 to a suitable DNA nucleotide of the oligo (for example an amino-modified C6 dT base to which NHS functionality has been added during manufacture). The 16 base long oligo tag 62 has a molecular weight of about 5,000 daltons, substantially less than the VEHICLES described in the embodiment shown in Figure 3, and therefore less expensive and also able to diffuse and bind BINDERS more rapidly in solution. Those skilled in the art will be able to design oligonucleotide tag sequences of reduced (or longer) length capable of specifically hybridizing with complementary sequences as required in the steps of Figure 4 C and D. TARGET and STANDARD oligonucleotide tags (e.g., 62 or 63) may be provided of lengths ranging from 4 to 30 bases, more preferably 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or 16-30 bases. Oligo tag sequences are designed according to well-known principles to maximize specific binding to a complementary sequence (e.g., 64 or 68) and dissociate at a reasonable (melting) temperature, while minimizing the potential to hybridize with other oligo sequences used in the workflow, or to form self-associations (e.g., intra-molecule hairpins or inter-molecular hybrids).

[0349] The product of the reaction of the oligo tag’s NHS and the peptide’s NH2 groups of Figure 4A is shown in Figure 4B, while the equivalent reaction for TARGET peptide 53 with TARGET tag 66 is shown in Figure 4 E and F. The STANDARD construct of Figure 4B is added to a sample digest (in this example a tryptic digest) whose sample peptides (including the TARGET) are prepared as TARGET -tag constructs (as in Figure 4F), thus forming a standardized sample digest. In order to preserve the quantitative relevance of the TARGET amount, it is desirable that the reaction of peptide 53 with VEHICLE 66 goes to completion (or if not, that the proportion of the TARGET peptide incorporated into construct of Figure 4F is consistent between samples).

[0350] At this point TARGET and STANDARD peptides are linked to, and thus identified by, oligo sequence tags of the respective TARGET and STANDARD tags (oligos 62 and 66 in Figure 4). Enrichment of the TARGET and STANDARD peptides using the cognate BINDER isolates these two peptides and their attached respective oligo tags (the peptides having identical structures but derived from different sources), preserving the TARGET-to- STANDARD ratio in the standardized sample. Subsequent to the enrichment step, after removal of unbound sample peptide constructs, the enriched bound TARGET and STANDARD constructs can be “completed” by hybridization and ligation to respective “secondary VEHICLES” as needed for various single molecule detection methods. Completed constructs shown in Figures 4 C and D (for the STANDARD peptide) and Figure 4 G and H (for the TARGET peptide) are particularly useful in the case of nanopore sequencing. In summary, short oligo tag 62 (which identifies peptide 52 as a STANDARD molecule) hybridizes with a complementary sequence 64 bringing the 3’ residue of the oligo 62 into proximity with the phosphorylated 5’ end of an oligo comprising an abasic region 36 (abasic backbone links are symbolized by “o”) and a following sequence 63. In some embodiments the 5’ terminus of the secondary VEHICLE comprises one or more bases before the abasic region (e.g., AA at the 5’ end of secondary VEHICLE oligo 36 in Figure 4) that are capable of hybridizing with complementary bases in oligo 70 (TT in oligo segment 70 of Figure 4D), and these bases are different in STANDARD and TARGET secondary VEHICLES so as to minimize potential hybridization of a secondary STANDARD oligo (e.g., 36 + 63) with a TARGET complementary sequence 70, and vice versa. The number of such hybridizing bases at the 5’ end of segments 36 may be selected to have an extended length that is less than the total length of linker 34 in order to avoid overlap of the peptide with these hybridizing bases when the complete construct passes through a nanopore. Similarly, the length of abasic region 36 is chosen so as to avoid overlap with the peptide 52 as it transits a nanopore, and oligo sequence 63 is designed so as to engage with a DNA motor and regulate pore transit of the peptide, allowing measurement of its current trace (“squiggle”). To optionally further reinforce the identification of the peptide as a STANDARD, oligo sequence 63 may be a unique sequence also indicating status as a STANDARD molecule (i.e., different from the sequence of oligo 67 in the TARGET VEHICLE construct) resulting in a “double-tag” (i.e., redundantly tagged) construct. An advantage of the double-tag approach is that each peptide is identified as STANDARD or TARGET by sequence information both before and after the peptide. In some embodiments, alternative linkage chemistries, including various click chemistry linkages as described elsewhere herein, are used to create a linkage between a peptide and a nucleic acid tag. Figure 5 shows an example in which a tag oligo prepared with a 3’ transcyclooctyne (TCO) group is reacted with a peptide derivatized at its n-terminus with methyltetrazine (Me-TZ) to yield an “in-line” construct (one in which the oligo and peptide sequences form a single continuous polymer). Figure 5A shows preparation of a TARGET construct using a tag sequence identifying the construct as a TARGET (labeled OLIGO- TARGET tag). In some embodiments, this modification (labeling peptides with the OLIGO- TARGET tag) is applied to all peptides in a sample digest (for example by derivatizing all digest peptide amino groups with NHS-tetrazine and reacting these with an excess of the reactive oligo comprising 3’ TCO), thereby ensuring that all molecules of the TARGET are taken into account in subsequent steps of an analytical workflow. Figure 5B shows preparation of a STANDARD construct using a tag sequence identifying the construct as a STANDARD (i.e., the OLIGO-STANDARD tag). In some embodiments, the STANDARD construct is prepared separately, for example using synthetic STANDARD peptide having the same (or generally similar) sequence as the TARGET, and is added to sample digests to serve as an internal quantitative standard (i.e., to standardize the sample digests). In some embodiments the physical properties of the TARGET and STANDARD oligo tags (e.g., length, base composition, and melting temperature) are selected to be similar so as to minimize the impact of their different sequences (necessary to distinguish then during peptide single molecule analysis) on relative binding to the cognate BINDER. Figure 5C shows TARGET and STANDARD constructs bound to cognate antibody BINDER molecules immobilized on a magnetic bead. Since the difference between TARGET and STANDARD is encoded by the oligo component, the peptide components can be chemically identical (same sequence) capable of identical interaction with the BINDER. In some embodiments, the peptide is linked to the oligo via its c-terminus (e.g., via a c-terminal carboxyl and the epsilon amino group of a c-term lysine) rather than the n-term as shown. In some embodiments the oligo is linked to the peptide via the 5’ end rather than the 3’ as shown. These and other functionally equivalent physical arrangements will be apparent to those skilled in the art.

[0351] In some embodiments, such as nanopore sequencing applications, that may require addition of molecular components such as oligonucleotides to both ends of a peptide, STANDARD constructs can be prepared in a stepwise process comprising 2 discrete attachment steps, thereby allowing different molecular components to be added at (or near) the peptide N-terminus and C-terminus, and establishing a consistent orientation of the peptide with respect to the overall construct (thereby avoiding the need to recognize a peptide in two different polarities). In some embodiments the 2-step digestion process described herein for preparing sample peptide libraries is used to generate STANDARD constructs. In this case, a synthetic peptide comprising the cognate TARGET sequence ending in Lys (i.e., a 2-amino peptide), and comprising an added n-terminal sequence ending in Arg (e.g., GSGR in the case of trypsin second cleavage, or any suitable peptide ending in the amino acid at whose c- terminal position a second proteolytic enzyme cleaves), can be processed in a series of steps similar to those used to process sample peptides as shown in Figure 6. For example, for a generic TARGET peptide sequence XXXXXXK, a synthetic STANDARD precursor can be generated with the sequence XXXRXXXXXXK. This peptide has both an n-terminal and a lysine epsilon amino group, both of which can be reacted with a suitable linkage reagent such as NHS-BCN, thus adding a click linker to both ends of the peptide, after which excess NHS- BCN can be removed. A suitably activated flag group (e.g., an oligonucleotide) can be linked to the peptide at this point using an appropriate click partner (e.g., azide to react with BCN) in order to identify the STANDARD (either generically using the same flag for all STANDARDS, or specifically, using a flag that identifies which STANDARD peptide sequence is involved). Cleavage with a second enzyme (e.g., trypsin) then reveals a fresh n-terminal amino group (in the peptide XXXXXXK) which is available to react with a second linkage reagent such as NHS-TCO, thereby activating the n-terminus for potential connection to a second tetrazine- activated oligo via the TCO-tetrazine click reaction. Use of different mutually orthogonal click groups at the two termini offers the ability to add distinct oligos to the two ends via the distinct click pair reactions, and to postpone one or both oligo additions until after BINDER enrichment of the TARGET and cognate STANDARD peptides. Likewise, one or both of the oligos may be coupled to the peptide prior to enrichment on a BINDER, provided that the same process is applied to the peptides of the digest (including TARGET peptides) prior to enrichment by a BINDER. It will be apparent to those skilled in the art that alternative linkage chemistries, including different click pairs or application of click pairs in alternate order, can be employed; that the steps of adding the first oligo for click linkage and second cleavage can be carried out in either order, and that the addition of the second linker group and / or second oligo may be delayed until after the BINDER enrichment step or omitted altogether when the single molecule sequencing and detection means requires an available n-terminal amino group (e.g., for Edman degradation).

[0352] An important advantage of carrying out BINDER enrichment after assembly of peptide-oligo constructs such as those shown in Figures 3, 4, 5 and 7 is that constructs taken forward for sequencing after enrichment are very likely to contain a peptide (otherwise they would not be bound by and recovered from the BINDER).

[0353] In some embodiments, a TARGET peptide, its cognate STANDARD and the cognate specific affinity BINDER reagent(s) (i.e., those molecules used together in the present invention to quantitate the TARGET peptide) may be developed together (co-evolved or co- optimized) to achieve optimal performance; i.e., through an iterative process comparing assay performance of various combinations of versions of the reagents, through a full-matrix comparison of all available variants of each, or by molecular engineering guided chemical knowledge and experimental results.

[0354] 6.4.8 Multi-level standards

[0355] In some embodiments, multiple distinguishable STANDARDS are provided for a single TARGET, and added in different amounts so as to establish a standard curve against which the TARGET can be quantitated. In such embodiments, the multiple STANDARDS may be distinguished by connection to distinct oligo sequence tags, or they may contain different and distinguishable structural modifications of the TARGET peptide (e.g., different amino acids added to its sequence). The different STANDARDS may be added to a sample digest in different amounts to standardize it: for example, three STANDARD versions (A, B, C) of a given TARGET peptide may be added to a digest in 0.1 : 1.0 : 10 relative amounts, thus generating a 3-point calibration curve. Multi-level STANDARDS increase the likelihood that at least one STANDARD will be present in an amount (and thus number of molecules) close to the amount of a TARGET in an unknown sample. Likewise, the pre-established ratio between different STANDARD version provides an internal check on the quantitative precision and linearity of a single molecule detection system. 6.5 SPECIFIC AFFINITY REAGENTS (BINDERS) TO CAPTURE AND ENRICH PEPTIDE TARGETS AND STANDARDS.

[0356] One or more specific affinity reagents, capable of binding the peptide TARGET and STANDARD specifically (i.e., while not binding a potentially vast number of other peptides that can be present in a sample digest) are used in some embodiments of the invention to capture the TARGET peptide and STANDARD prior to single molecule detection. We refer to such a reagent generically as a BINDER, and include within that term not only canonical antibodies such as IgG, but also numerous types of proteins and other macromolecules (e.g., aptamers) known in the art to be able to bind to particular peptide sequences with specific affinity.

[0357] Experience with the SISCAPA technology (3, 9, 13-15) has shown that antibodies, preferably monoclonal antibodies, can be developed that bind and capture a specific low abundance tryptic peptide from the digest of a very complex sample such as human blood plasma (which may contain 250,000 distinct peptides, some at very high abundance), and thereby enrich the peptide substantially (e.g., more than 10,000-fold). A variety of types of biologically derived antibodies (e.g., polyclonal, monoclonal and oligoclonal antibodies derived from mice, rabbits, humans, camelids and other species), molecules derived from antibodies by molecular biology techniques (e.g., antibodies selected from libraries using phage display and other techniques), aptamers, and other molecular constructs can be created to achieve the purpose of specific peptide binding - all of these are included within the term BINDER as used herein.

[0358] 6.5.1 Antibody BINDERS

[0359] Many methods exist for generating antibodies to a peptide in animals. For example, a synthetic peptide having the TARGET peptide sequence can be coupled to a carrier protein (e.g., keyhole limpet hemocyanin: KLH) and used to immunize an animal (such as a rabbit, mouse, chicken, goat, camelid or sheep) by one of the known protocols that efficiently generate anti-peptide antibodies. For convenience, the peptide used for immunization and antibody purification may contain additional c-terminal or n-terminal residues (e.g., cysteine) added to the TARGET peptide sequence. The resulting extended TARGET peptide can be conveniently coupled to carrier KLH that has been previously reacted with a heterobifunctional reagent such that multiple SH-reactive groups are attached to the carrier. In classical immunization with the peptide (now constituting a hapten on the carrier protein), a polyclonal antiserum can be produced containing antibodies directed to the peptide, to the carrier, and to other non-specific epitopes.

[0360] Specific polyclonal anti-peptide antibodies can be prepared from an immunized animal’s serum by affinity purification on a column containing tightly-bound peptide. Such a column can be easily prepared by reacting an aliquot of synthetic TARGET peptide made with a cysteine residue added to one end with a thiol-reactive solid support. Crude antiserum can be applied to this column, which is then washed and finally exposed to 10% acetic acid (or other elution buffer of low pH, high pH, or high chaotrope concentration) to specifically elute anti-peptide antibodies. These antibodies are neutralized or separated from the elution buffer (to prevent denaturation), and the column is recycled to physiological conditions for application of more antiserum if needed.

[0361] In some embodiments, one of a variety of methods known in the art (e.g., B-cell cloning, hybridoma generation, recombinant expression, etc.) is used to generate candidate clonal antibody proteins, genes or gene sequences (from the cells, genes, or proteins of an immunized animal, or from natural or artificial protein binder libraries such as phage display libraries) that can be screened to select a monoclonal antibody (or other BINDER molecule) with the ability to enrich the TARGET peptide and cognate STANDARD from a complex peptide mixture (e.g., a sample digest) under a specified set of solution conditions. Such screening can be carried out using the method of the invention (i.e., screening in the assay of ultimate use), or by alternative methods such as the SISCAPA method with MS detection. Monoclonal antibodies, particularly those produced by recombinant methods, have the advantages of homogeneity, superior performance, scalable production and longevity compared to polyclonal mixtures.

[0362] In some embodiments a preferred method of selecting a monoclonal (homogeneous) anti-peptide antibody BINDER for use in the invention includes testing whether a candidate antibody (e.g., product of a clone) binds the TARGET peptide and STANDARD equally, and selecting for use the antibody product of a clone or clones that bind these peptides most equally (i.e., with least bias towards one or the other, and thus capable of capturing both from a mixture without changing the ratio between them). Preserving the ratio of TARGET to STANDARD unchanged during capture and enrichment by the BINDER is desirable since the invention involves measuring this ratio (by counting TARGET peptide and STANDARD molecules) to calculate the amount of TARGET peptide (and hence target protein) in the original sample. In the event that BINDERS are not found that preserve the TARGET to STANDARD ratio precisely, some embodiments make use of BINDERS that exhibit differential, but reproducible, binding, thus allowing correction of the ratio as described above.

[0363] 6.5.2 Multi-epitope BINDERS

[0364] In some embodiments multiple BINDERS are used to enrich a single peptide. In some embodiments, TARGET peptides are sufficiently long to include multiple epitopes: a typical antibody linear epitope is 4-6 amino acids long, while TARGET peptides may be 6-30 amino acids in length - those of length 12-30 amino acids have a high likelihood of comprising 2 or more non-overlapping epitopes. In some embodiments, multiple BINDERS targeting linear, non-overlapping epitopes in a peptide can be generated and used to increase enrichment specificity. In some embodiments monoclonal antibodies to multiple linear, non-overlapping epitopes of a peptide are made using conventional hybridoma or other cloning techniques to select among clones created by immunization of an animal with peptide and / or derivatives of it. In some embodiments, antibodies to multiple linear, non-overlapping epitopes of a peptide are selected from libraries such as naive or immunized phage display libraries of single-chain antibodies. In some embodiments, aptamers to multiple linear, non-overlapping epitopes of a peptide are selected from libraries or evolved using iterative selection approaches well-known in the art. In some embodiments, multiple BINDERS of different types (e.g., polyclonal or monoclonal antibodies, aptamers, etc.) targeting linear, non-overlapping epitopes are used together for increased affinity and / or specificity.

[0365] In many cases, the cost involved in creation of multiple monoclonal antibodies or aptamers for each TARGET peptide can present a practical barrier to such an approach. Some embodiments therefore make use of polyclonal antibodies purified from sera of immunized animals, or, alternatively, oligoclonal mixtures of antibody-like molecules extracted from large libraries (e.g., naive or immunized phage display libraries). In some embodiments methods well known in the art are used for affinity purification of multiple distinct antibody specificities from such polyclonal antisera using multiple affinity media, each comprising a linear peptide subsequence of a TARGET peptide sequence.

[0366] In some embodiments, multiple BINDERS with affinities for distinct, ideally non- overlapping, epitopes on a TARGET peptide are simultaneously affinity purified by binding to synthetic TARGET peptide. In this approach, a TARGET peptide with multiple epitopes binds multiple BINDERS from a mixture until each epitope is saturated with BINDER, thereby establishing a balanced mixture of BINDERS to the epitopes. Figure 8B shows 3 BINDERS (I, II and III) bound to different linear epitopes (1, 2, 3) on a single peptide molecule (the peptide shown in Fig 8A). In some embodiments, this is achieved by affinity purification of BINDERS from a polyclonal antiserum (or pool of antisera) generated in response to immunization with the peptide, or multiple fragments of it. In some embodiments, this is achieved by capture of a mixture of BINDERS from a variety of sources selected to bind to the peptide, or fragments of it. By saturating the TARGET peptide epitopes with BINDERS competing to bind to various epitopes, a population of BINDERS is produced that covers the peptide (i.e., has a BINDER bound to all, or at least a substantial fraction, of the epitopes present on each peptide molecule). The BINDERS are expected to have varying affinities and specificities, some or most of which may be individually insufficient to effectively capture the peptide from a sample digest. However, in some embodiments, a combination of these BINDERS can be linked together covalently (e.g., using bi-functional chemical crosslinkers or click connectors) or non-covalently (e.g., by reaction of biotin-labeled BINDERS with multivalent streptavidin) into a multivalent (e.g., bi-specific, tri-specific, or quadra-specific) BINDER (Fig 8C), that can be eluted from the peptide and subsequently used for affinity enrichment of TARGET and cognate STANDARD from standardized sample digests. This novel approach makes use of a multi-epitope peptide as a “scaffold” or “template” on which a plurality of BINDERS to short epitopes are assembled and then linked to form a larger multi- epitope BINDER. In some embodiments a multi-epitope (or multi-valent) BINDER is immobilized (e.g., on magnetic beads, a column, a surface, etc.) and used to capture and enrich TARGET and STANDARD peptides and peptide constructs according to the invention.

[0367] Such a multi-valent BINDER can bind the peptide with much higher affinity than any of the individual BINDERS. This effect is well known in the art as “avidity”: the increased binding efficacy of a multivalent BINDER compared to a monovalent BINDER. Natural antibodies exploit this effect, comprising 2 binding sites in IgG and 5 binding sites in IgM. In some embodiments, this avidity effect is exploited by crosslinking the individual epitope BINDERS while in proximity to one another (in situ in Figure 8C) to create a multi-BINDER construct as shown in Figure 8D. In some embodiments, a similar approach is carried out using individual monoclonal BINDERS, either crosslinked together or expressed in a combined recombinant product similar to well-known “bi-specific” or “tri-specific” therapeutic antibody constructs.

[0368] In some embodiments, multiple BINDERS with affinities for distinct, ideally non- overlapping, epitopes on a TARGET peptide are separately affinity purified by binding one after another to a series of immobilized synthetic TARGET peptides. In some embodiments, multiple BINDERS with affinities for distinct, ideally non-overlapping, epitopes on a TARGET peptide are used to sequentially affinity purify TARGETS and cognate STANDARDS from a standardized digest: TARGETS and cognate STANDARDS that bind to (and are subsequently eluted from) each successive BINDER specific to one of multiple TARGET epitopes are rendered purer than the set of TARGETS and cognate STANDARDS that bind to and are eluted from a BINDER to a single epitope.

[0369] In some embodiments, the higher peptide sequence specificity obtainable with multi- epitope BINDERS (through their interaction with a larger number of amino acids in the peptide portions of TARGETS and STANDARDS) provides an important addition to the overall specificity of a range of single molecule detection technologies. By increasing the enrichment discrimination between authentic pre-selected TARGET / STANDARD peptide molecules and the multitude of potential similar sequence “ off-target’ ’ peptides present in digests of complex samples, potential assay interferences (particularly false-positive detection of TARGET -like peptides) is reduced for all single molecule detection methods.

[0370] Multi-epitope BINDERS are described in more detail in U.S Provisional Patent Application No. 63 / 381,722, incorporated herein by reference in its entirety.

[0371] 6.5.3 Enrichment of TARGETS and STANDARDS from standardized sample digests

[0372] The invention makes use of a specific enrichment step to enrich both TARGET peptide and cognate STANDARD from a sample digest, thereby creating an “enriched standardized digest sample” (or “enriched standardized sample digest”). In some embodiments the enrichment step is carried out by a cognate affinity capture reagent (BINDER) to which the two peptides bind equivalently, such that the ratio of their amounts after enrichment is the same or nearly the same as before enrichment. In some embodiments, the TARGET and STANDARD peptides bind to the cognate affinity capture reagent with identical affinity and kinetics, preserving the ratio between them exactly. In some embodiments the TARGET to STANDARD ratio after enrichment is within 2%, within 5%, within 10%, within 20%, or within 30% of the ratio before enrichment. In some embodiments, enrichment using a BINDER results in a change in the TARGET to STANDARD ratio, which change is consistent across a range of samples and assay replicates. In this case, prior knowledge of the factor by which the ratio is changed by enrichment (established by measurements of TARGET to STANDARD ratios in a sample or digest before and after enrichment) allows correction of the ratio observed after enrichment to yield the correct relative amounts of TARGET and STANDARD in the sample digest.

[0373] In some embodiments, a homogeneous cognate affinity capture reagent (BINDER, described below) is selected from a plurality of alternative BINDERS for its ability to bind the TARGET and STANDARD peptides equivalently (or with a consistent ratio shift). Alternatively, in some embodiments, TARGET and STANDARD peptides are selected from a plurality of alternatives to bind to a cognate BINDER equivalently. In some embodiments, TARGET peptide, STANDARD peptide, and cognate BINDER are each selected from a plurality of alternatives so as to maximize the property of equivalent (or consistent) TARGET and STANDARD peptide binding.

[0374] Conservation (or correctability, as described above) of the ratio between TARGET and STANDARD peptides through enrichment means that a measurement of the TARGET peptide to STANDARD ratio after enrichment provides an accurate measure of TARGET peptide amount, even if only a fraction of the total TARGET and STANDARD is captured in the enrichment step. For example, the enrichment process might capture 100%, or 90% or 10% or 1% or 0.1% or 0.01% or 0.001% of the TARGET and STANDARD peptides present in a sample digest, but in each case the ratio of TARGET peptide to STANDARD molecules would remain the same, and thus a measurement of the TARGET peptide to STANDARD ratio would provide the same answer, which would be equal to (or correctable to) the ratio present in the original standardized sample.

[0375] The proportion of a peptide captured by the enrichment step is therefore an adjustable feature of the invention, which can be used to capture more of one TARGET peptide & its cognate STANDARD than another. Such adjustments make it possible to use the enrichment step to capture most or all of a low abundance TARGET peptide, while capturing only a small fraction of the molecules of a high abundance TARGET peptide - with the result that the difference in absolute molar amounts of the two TARGET peptides can be significantly reduced by this differential enrichment. This use of the enrichment step to bring multiple TARGET peptides to similar abundances (while preserving the TARGET peptide to STANDARD ratios that encode TARGET peptide amount in the original sample) is referred to as “stoichiometric flattening” or “equalization”, and provides a means by which amounts of molecules with high and low abundances in the original sample can be measured using a measurement technology with limited dynamic range (e.g., single molecule counting methods). The flattening approach (differential enrichment of different TARGET / STANDARD cognate pairs) generates a “flattened enriched standardized sample digest sample”.

[0376] 6.5.4 BINDER affinity

[0377] In some embodiments the affinity of the BINDERS most useful for enrichment is in the range of 0.01 to 10 nanomolar, more particularly with a preferred half-ofif-time (reciprocal of the off-rate) of at least several minutes, or more preferably 10-15 minutes. Off-rate is particularly important since it governs the length of time that unbound materials, including non-TARGET peptides, can be washed away (e.g., using conventional manual and automated workflow steps such as magnetic bead manipulation in 96-well plates) while retaining the TARGET and STANDARD peptides on the BINDER. Higher affinity BINDERS are typically required to enrich lower abundance TARGETS, i.e., to capture peptides present in a digest at low concentration.

[0378] In some embodiments, specific solution conditions, or changes in solution conditions, are employed in a sample preparation workflow to preferentially dissociate less-tightly-bound peptides while retaining the correct BINDER cognate TARGETS and STANDARDS. In some embodiments bound peptides are exposed to increasing chaotrope or denaturant concentrations, increasingly acidic or basic solution pH, increasing salt concentrations, and / or increasing temperature to dissociate less-tightly-bound peptides prior to final elution of the enriched TARGETS and STANDARDS.

[0379] In some embodiments (e.g., single molecule imaging technologies) in which labeled BINDERS are used to detect and identify specific TARGET peptide sequences, BINDER specificity may be more important than affinity, and BINDER selection methods correspondingly adapted.

[0380] 6.5.5 Dissociation from BINDER

[0381] In some embodiments, a preferred property of a BINDER is the ability to release bound peptides rapidly at a desired point in a workflow, as a result of a change in solution conditions - for example as a result of a change in pH (e.g., pH 3.0 or pH 9.5), addition of a chaotrope (e.g., ammonium thiocyanate or KC1) or organic solvent (e.g., 50% acetonitrile in water), increase in temperature, or the application of an electrical field (as in electrophoresis).

[0382] In some embodiments, BINDERS are selected that tightly and specifically bind the cognate TARGET and STANDARD peptides from a digest, and then release these peptides only when in close proximity to a site at which sequence-sensitive single molecule detection can occur (e.g., a nanopore, or an immobilization site on a support), thus maintaining the peptides in concentrated form and reducing losses due to diffusion. Nanopore sequencing (53) typically relies on the presence of a concentrated salt solution (e.g., 0.4M KC1) to provide sufficient charge carriers to create a measurable open channel current through the nanopore (typically 20-200 pA). In some embodiments BINDER reagents are selected that release their TARGET and STANDARD peptide cargo when exposed to such salt conditions, i.e., when the antibodies are placed in the solution present on the cis side of a nanopore sequencing device, or when exposed to a high concentration of salt in a salt gradient in which salt concentration increases closer to a nanopore. This feature of a selected BINDER (i.e., releasing bound peptides in a high-salt environment) is advantageous in that peptides can be retained on a physically manipulatable solid support (e.g., magnetic beads) until they are in a sequencing chamber itself (or in a region of that chamber nearest to a nanopore), thereby minimizing potential losses and dilution that could occur if peptides were eluted elsewhere and later transported into the sequencing chamber. In some embodiments, chaotropic anions such as SCN are incorporated into the solutions of a nanopore cis compartment (or both compartments) in addition to or in place of Cl anions conventionally used, in order to facilitate release of peptides from a BINDER. It will be evident to those skilled in the art that a range of chaotropic anions or cations can be used to effect peptide release from BINDER over a range of concentrations suitable for optimization in particular device configurations.

[0383] In some embodiments using one or more alternative influences to release bound peptides from a BINDER without a pH change, the potential deleterious effects of acid elution (as practiced in conventional affinity enrichment systems including SISCAP A, e.g., pH 2.0- 3.5) on some protein nanopores, or on components of other single molecule detection systems, can be avoided. Acid elution is used in SISCAPA in order to avoid introduction of salts, since salts interfere significantly with detection by mass spectrometry. The use of high salt in nanopore sequencing at near-neutral pH contrasts with the use of acid elution and low salt in mass spectrometry-based detection systems, and this difference suggests that different BINDERS with different elution characteristics will be preferred in the respective peptide detection methods.

[0384] In some embodiments where enriched TARGET and STANDARD constructs are to be immobilized on a support, the constructs are delivered into proximity with the support while they are bound to the cognate BINDER (e.g., on easily manipulatable magnetic beads). BINDER-bound constructs are present at a very high effective local concentration, and may be conveniently moved from one environment to another without loss. This feature is an important advantage of the BINDER enrichment step when applied to low abundance peptides and their detection by single molecule counting.

[0385] 6.5.6 BINDER immobilization

[0386] In some embodiments, a BINDER with specific affinity for the TARGET peptide and STANDARD may be immobilized on a solid support in order to facilitate separation of the antibody and its bound peptide and / or peptide construct cargo from a complex sample digest, to wash away unbound molecules, to concentrate bound peptides, and to deliver bound peptides to a site where they are available for sequencing. Typical solid supports used for this purpose include magnetic beads (allowing collection of beads from a liquid suspension by magnetic force) or a porous column (e.g., an affinity column) through which liquids may be pumped. In some embodiments, the BINDER is immobilized on commercially available protein G- derivatized magnetic beads (Dynabeads G; Thermo Fisher) and optionally crosslinked covalently with dimethyl pimelimidate (DMP) according to the manufacturer’s instructions. In an alternative preferred embodiment, the antibody is immobilized on tosyl-activated Dynabead magnetic beads. In a further alternative embodiment, the anti-peptide antibody can be immobilized on solid phase chromatography media (e.g., POROS G resin) packed in a column and crosslinked using DMP. Such a column can bind the TARGET peptide specifically from a peptide mixture (e.g., a tryptic digest of serum or plasma) and, following a wash step, release the TARGET peptide under elution conditions.

[0387] 6.5.7 BINDER homogeneity

[0388] In some embodiments, e.g., those using a homogeneous cognate affinity capture reagent (e.g., a monoclonal antibody BINDER, wherein all or nearly all molecules have the same sequence), it is expected that the ratio of TARGET and STANDARD peptides is not affected by the degree of saturation of the BINDER binding sites by the peptides at equilibrium (particularly at low saturation). Inhomogeneous affinity capture reagents are difficult to characterize in detail, and can contain variants that bind one or the other of the TARGET and STANDARD peptides more strongly. Thus, saturation of one variant could be followed by binding to another, potentially lower affinity, variant that has different relative affinities for the TARGET and STANDARD peptides, resulting in a change in the bound ratio as a function of the amount bound. For this reason, homogeneous (typically clonal or sequence-defined) BINDERS are typically preferred: e.g., monoclonal antibodies or sequence-defined aptamers.

[0389] 6.5.8 Chemical modification of peptides while bound to BINDER

[0390] In some embodiments, chemical or enzymatic reactions for the purpose of modifying a TARGET (or STANDARD) peptide are carried out in solution, and in some embodiments one or more reactions are carried out while a peptide or peptide construct is bound to a BINDER, which may or may not itself be bound to a solid support. In some embodiments, one or more reactions are carried out while a peptide is bound to a BINDER linked to a solid support, thus allowing the peptide to be contacted with reagents, and removed from contact, by physical movement of the support between liquids (e.g., by removal of magnetic beads carrying BINDER and bound peptides from liquid in one vessel and deposition of the beads in a different vessel where they are exposed to a different reagent), or equivalently by movement of liquids in contact with the support (e.g., by pumping one reagent and then a second reagent over a porous column, or magnetic bead mass, to which BINDER and its peptide cargo are bound). In addition to contact with one or more reagents required for execution of a sequence of reactions, manipulation of peptides on a support allows the peptide to be washed free of a reagent by exposure to a wash solvent prior to contact with a subsequent reagent. Movement of peptides between liquids by movement of a BINDER or support to which they are bound reduces or eliminates the need for purification or concentration of intermediate peptide forms created during a sequence of one or more chemical reactions. In some embodiments, peptides are bound to a solid support by means other than interaction with a specific BINDER, e.g., by binding of peptides to a generic support such as a reversed phase support (e.g., C18) or an ion exchange support.

[0391] In some embodiments, use is made of amino groups present at peptide amino termini and on lysine side chains for chemical linkage of a peptide to other molecules (e.g., oligo and other polymers that, with peptides, form constructs amenable to sequence-sensitive single molecule detection) while a peptide is bound to a BINDER. In order to eliminate competing side reactions with amino groups present in specific affinity reagents (e.g., lysines and n- terminal amino groups of anti-peptide antibody BINDERs) used in the invention, the invention provides for the optional elimination of some or all of these BINDER amino groups by chemical blockage (e.g., by reductive methylation, by PEGylation using commercially- available NHS-PEG or other reagents, conversion of lysine residues to homoarginine by treatment with O-methylisourea, or other chemical modifications known in the art), by protein engineering (e.g., by replacing some or all lysines in a recombinant antibody sequence with arginines or other amino acids), or various other means. Specific affinity reagents to be used in such embodiments may be selected so as not to contain any lysine residues in the TARGET peptide binding site, since these residues would likely be blocked along with other lysines, potentially leading to a loss of binding activity. Non-protein BINDERS, such as DNA and RNA aptamers and other similar molecules, may contain no amino groups to begin with, eliminating the need to block these prior to process aimed at modifying peptide amino groups. The elimination of BINDER amino groups that could participate in side reactions has the effect of avoiding waste of expensive reagents used in amino group modifications of TARGET and STANDARD peptides, including use in creating concatenated constructs of these.

[0392] In some embodiments, blockage (e.g., by PEGylation) of many or all of the amino groups on BINDERS, and on any other proteins present on the capturing support (e.g., Protein A or Protein G used to guide antibody immobilization on solid supports such as Dynabeads G magnetic beads), can also have the advantageous effect of rendering the BINDER more stable, and thus less liable to degradation by heat, by proteases, or by exposure to complex samples and sample digests. In the case of a Protein A or G coated magnetic bead, for example, it is advantageous in some embodiments to first react the antibody BINDER with the Protein A or G on the bead, then chemically cross-link the BINDER to the Protein A or G on the bead, then PEGylate some or all of the remaining protein amino groups on the bead. Such modifications can also alter the net charge on proteins and on beads carrying them towards greater negative charge overall, since typically about half the positive charges on a protein are arginine and half lysine (the latter of which would be blocked by blockage of amino groups). Since the amount of negative charge (largely attributable to glutamic and aspartic acids) would be unaffected by amino group blockage, the overall decrease in positive charges by -50% will shift the net charge on the BINDERS, and on a bead coated with BINDERS, towards the negative. Nanopore sequencing devices are typically operated with a negative electrode in the cis compartment (where the input molecules to be sequenced are added) and a positive electrode in the trans compartment: this polarity induces an oligo, which is strongly negatively charged on account of its sugar-phosphate backbone, to migrate towards and through the pore to initiate sequencing. In some embodiments this polarity also serves to move a negatively charged bead towards the pore, contributing towards the goal of delivering peptide-oligo constructs in close proximity to the pore.

[0393] 6.6 BARCODES

[0394] In some embodiments the methods used for single molecule detection have the capability to detect very large numbers of molecules (e.g., IO10in (54) far exceeding the requirements for quantitative measurement of a modest number of peptides in one sample. In order to make effective use of the analytical capacity of such platforms, and the consequent reduction in analytical cost, some embodiments connect sample-specific labels (“barcodes”) to the TARGET and STANDARD peptides present in a sample digest, or enriched from a sample digest by a BINDER, allowing the TARGET and STANDARD peptides from multiple samples to be combined prior to peptide detection (i.e., multiplexed), and afterwards de- multiplexed to associate them with the correct original samples. DNA provides an ideal medium for implementation of such barcodes since, as essentially a digital medium, it is easy to synthesize, cut, ligate, copy, and detect by both sequencing and hybridization. Alternative barcode polymers can be employed, such as peptides and synthetic chemical polymers, although these may be significantly more difficult to generate, manipulate and detect than oligonucleotides.

[0395] As described above, many sample barcoding systems have been developed using sets of distinct DNA barcodes to identify nucleic acid molecules derived from different samples prior to sequencing, or to facilitate optical readout of individual nucleic acid molecules in imaging systems. In some embodiments, sample barcodes with identical or very similar base composition but distinguishable sequences are preferred in order to minimize differences in physical properties between constructs on account of barcode properties.

[0396] In some embodiments the identity of samples from which single molecule constructs according to the invention are derived is encoded using nucleic acid (e.g., DNA or another sequenceable polymer having multiple distinguishable subunits) barcodes. In some embodiments sample barcodes are appended or linked to TARGET and STANDARD constructs prior to enrichment by cognate BINDERS. In some embodiments sample barcodes are appended or linked to TARGET and STANDARD constructs after enrichment by cognate BINDERS, in which case smaller amounts of the DNA barcodes are required.

[0397] Figures 9 and 10 illustrate schematically a 2-level encoding scheme used in some embodiments. In each sample digest, a specific peptide (here labeled Peptide- A) is linked to a DNA sequence tag (labeled OLIGO-TARGET) identifying it as a sample-derived TARGET molecule. The cognate internal standard formed by linkage of a synthetic version of Peptide- A with a distinct DNA sequence tag (labeled OLIGO-STANDARD) is added to the digest, creating a standardized digest (standardized with respect to Peptide-A). Either before, or more efficiently after, enrichment of Peptide-A TARGET and STANDARD constructs using a BINDER, sample barcodes comprising a plurality of modules (Codes) are linked to these constructs using conventional methods that may include ligation to the TARGET / STANDARD tag, chemical linkage (e.g., using click chemistry), non-covalent means (e.g., biotin on one oligo and streptavidin on the other), or a variety of other linkage means known in the art. Alternatively, the sample barcodes can be linked to a site on the peptide different from the site at which the TARGET / STANDARD tag is connected.

[0398] The scheme for sample barcoding shown in Figure 9 provides a construct compatible with a variety of single molecule detection methods, as described below. In this example, barcode modules at positions 1, 2, 3 and 4 are used to encode “bits” in a 10-bit binary sample code. In such a code, if all bits were readable at once, 10 bits could conventionally encode 210= 1,024 samples. Given that each DNA base is 1 of 4 alternatives (2 bits of information), 10 bits of information could theoretically be encoded in a short sequence of 5 bases. However, all the methods envisioned for reading the sample barcode in a single molecule detection system are subject to error, and avoiding sample misassignment errors is a high priority in many applications (e.g., clinical). A preferred approach is therefore to add redundancy to the sample code. In some embodiments this is done by providing a unique sequence module comprising multiple bases (e.g., 4 to 30 bases depending on the preferred readout method) corresponding to each of the bits in the desired sample code space (number of samples to be identified). As a further measure against sample assignment errors, the error detection and correction methods of Hamming can be used, and in the case of a 10-bit code, Hamming extended parity error detection involves the addition of 4 parity bits to the 10-bit code, resulting in a total of 14 bits of information. Such a 10+4 = 14-bit code is capable of detecting and correcting any 1 -bit error, and detecting but not correcting 2 -bit errors. To implement such an approach in a manner that economizes on the total length of DNA that must be “read” to obtain the sample code, the example of Figure 9 simplifies this coding scheme to use only those code values having 3 or 4 bits set to a value of 1, which reduces the sample coding space to 105 samples that can be identified with very high accuracy, but reduces the total number of DNA modules that need to be in any one sample code. Thus 4 modules are included in any sample code, selected from among 14 different DNA sequences selected using computational and experimental methods well known in the art for minimal likelihood of confusion during readout. A mistaken read on any one of these modules can be corrected by the coding scheme, and mistakes in 2 modules can be detected (but not corrected). Those skilled in the art will recognize that many alternate coding schemes exist, with greater or lesser numbers of bits, of larger of smaller numbers of identifiable samples, and of great or lesser numbers of bases in each DNA module.

[0399] 6.6.1 DNA sample barcodes used with platforms that incorporate DNA sequencing.

[0400] Several approaches to peptide single molecule detection include the ability to read nucleic acid sequences interspersed with peptide sequence (e.g., current DNA sequencing nanopore platforms) or else together with peptide sequence that has been reverse-translated into DNA (e.g., reverse translation platforms). In some embodiments, modules of 2, 3, 4, 5, 6, 7, 8, 9, 10, 11,12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30 or more bases are used.

[0401] 6.6.1.1 Nanopore sequencing

[0402] DNA sample barcodes and TARGET vs STANDARD tags linked to peptides can be read directly by passage through a suitable nanopore sequencing system. At the current state of the art represented by the commercial MinlON device, the accuracy of individual basecalls can be greater than 99%, and therefore the accuracy with which one of a small set of sequences (designed to be distinct) can be recognized is high. In some embodiments, the sample barcode (e.g., a binary number identifying the sample, an alphanumeric code taken from a physical sample label, or any type of computer encodable sample identification) can be encoded directly (2 -bits per base) or with redundancy in the form of multiple bases per code bit, additional parity bits (including error detection and correction), or any information representation scheme that can be encoded in DNA or another nanopore-readable polymer with 2 or more distinguishable units. The association of a code with an individual peptide molecule is accomplished through the covalent linkage between the two that is established in the peptide library preparation workflow of the invention.

[0403] 6.6.1.2 Reverse translation

[0404] Methods of single molecule sequencing by some form of reverse translation have been described (US 2021 / 0302431 and (34)), and these typically include the ability to copy a DNA sequence from an affinity reagent capable of recognizing a terminal amino acid (or amino acids) to a “root” oligo attached to, or in proximity to, the peptide being reverse-translated, which is extended with a DNA code identifying the amino acid at each degradation cycle. In some embodiments, such a root oligo linked to a peptide comprises a TARGET / STANDARD tag (identifying which version of a peptide it is attached to) and optionally a sample barcode. After the required number of decoding and degradation steps, the root oligo is prepared for conventional high-throughput DNA sequencing, generating a sequence comprising the peptide sequence (or a representation of it), its identity as a TARGET or STANDARD, and its sample of origin.

[0405] 6.6.2 DNA sample barcodes used with affinity reagent imaging platforms

[0406] In some embodiments, e.g., those making use of single molecule imaging (30-32), DNA sample barcodes can be detected by sequential hybridization with labeled oligos complementary to the sample barcodes. Complementary oligos can be labeled with a variety of fluorescent or colored dyes, with quantum dots or other optically detectable nanoparticles, with enzymes capable of generating a localized signal (e.g., luminescence), or a variety of other compositions known in the art for the generation of a spatially localized extemally- detectable signal. In some embodiments, a set of barcode sequences is used that are designed to have high specificity (minimal cross-hybridization of one barcode with the probes complementary to the other barcodes). The lengths of the barcode sequence modules generally impact the specificity with which they are recognized by complementary probes, the kinetics with which they bind and the temperature at which they can be removed after being read (i.e., analogous to the “melting temperature”). In some embodiments, modules of 8, 9, 10, 11,12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30 or more bases are used.

[0407] Since imaging approaches typically do not reliably establish the order of individual barcodes present in a construct molecule (this being at or beyond the resolution limits of conventional imaging systems), probing a single molecule construct with a set of probes complementary to the set of barcodes yields a series of binary (i.e., the probe binds or does not bind) results that can be considered a binary code.

[0408] In some embodiments, a set of N distinct barcodes is used, where N is a number required to encode at least the number of samples to be pooled (i.e., 2N> number of samples). For example, 127 different samples could be uniquely identified by detecting the presence or absence of 7 different barcodes (27=128, yielding 127 unique barcodes, excluding 0 - the absence of all barcodes). To make use of this scheme would require that peptide:oligo constructs include up to all 7 of the barcodes.

[0409] In some embodiments, a larger set of barcodes is used, but constructs need include only a limited number of these. For example, if 11 distinct barcodes are used, but only 4 or fewer of these are included in any given construct, several hundred different samples can be uniquely identified. Use of fewer barcodes in a construct is advantageous since it reduces the length and cost of the sample barcoding oligos required: individual distinct barcodes may be 20-30 bases long.

[0410] In some embodiments, a further improvement in sample barcoding is provided by the use of Hamming codes. For example, in a coding scheme with 10 data bits and 4 parity bits (14 bits total, here corresponding to 14 distinct DNA barcodes), and using only 3 or 4 of these barcodes in any individual peptide construct, it is possible, using the features of the Hamming scheme, to identify and correct any single error in detection of any of the individual barcodes. Error detection is of great value in preventing mis-attribution of molecules to the wrong sample, which could result in erroneous quantitative results derived from errors in the counts of TARGET and STANDARD molecules in a sample digest.

[0411] Figure 10 illustrates the process of decoding an example construct in 16 cycles using a Hamming code: in each cycle one of 16 oligo probes complementary to one of the DNA codes is applied, detected when present, and removed. In this example the first 2 cycles involve probes that determine whether the peptide is a TARGET or a STANDARD (either one probe binds or the other). In the remaining 14 cycles, 14 probes complementary to each of the 14 DNA sequence modules described above are successively applied, detected when present, and removed. The resultis a 14-bit binary number that is capable of identifying one of 105 different samples with single (1-bit) error correction and double (2 -bit) error detection (referred to as “SECDED” in the art). Separately, but as part of the same decoding process, one or more recognition reagents capable of characterizing the peptide are applied in additional cycles to establish the identity of the peptide.

[0412] 7 GENERAL METHOD OF ANALYSIS. Some embodiments of the invention comprise a series of steps to transform a protein- containing sample into an enriched standardized digest sample, or a flattened enriched standardized digest sample, prior to sequence-sensitive single molecule detection and counting. The invention is equally applicable to protein samples from sources such as blood, blood plasma and blood serum, as well as other sources, such as tissue homogenates, animal, plant or microbial samples, other body fluids, environmental samples and the like.

[0413] An important feature of the invention is its generality, allowing the design of similar protocols and using similar reagents and equipment to prepare peptide libraries suitable for analysis of a wide range of different proteins in a variety single molecule detection systems. To accomplish this the invention makes use of specific features of peptides generated by particular enzyme cleavages; features provided by multiple click chemistry pairs; requirements for specific peptide and oligo orientation; multiple levels of barcoding; and detailed control of the capture and enrichment of peptide:oligo constructs by specific affinity reagents.

[0414] In some embodiments, steps of the general method may be carried out in a different order than that outlined below. Nevertheless the invention requires that steps of sample digestion to peptides, generation of TARGET constructs from TARGET peptides, and addition of STANDARD constructs to the digest (thus creating a standardized digest) must precede enrichment of TARGET constructs and STANDARD constructs from the standardized digest using BINDERS.

[0415] 7.1 DIGESTION OF SAMPLE.

[0416] In some embodiments a general approach for sequence-based protein quantitation involves digesting sample proteins (e.g., with trypsin) into peptides. In order to improve the completeness of digestion, disulfide bonds may be broken and proteins denatured to disrupt secondary and tertiary structure. Samples can be any kind of protein-containing sample without limitation, including body fluids, tissues, tissue lysates, tissue extracts, bacterial, fungal, animal and plant samples, recombinant proteins including protein drugs, food products, and the like.

[0417] In some embodiments, preparation of proteolytic peptides from a complex sample is carried out by a series of reagent addition steps which may include: denaturing a protein sample (e.g., with detergents such as deoxycholate or CHAPS, organic solvents, urea or guanidine HC1), reducing the disulfide bonds in the proteins (e.g., with tris(2- carboxyethyl)phosphine (TCEP), dithiothreitol or mercaptoethanol), alkylating the cysteines (e.g., by addition of iodoacetamide, or iodoacetic acid, which react with the free -SH group of cysteine preventing reformation of disulfide bonds), quenching excess iodoacetamide by addition of more dithiothreitol or mercaptoethanol, and (after removal or dilution of the denaturant) addition of the selected proteolytic enzyme (e.g. trypsin), followed by incubation to allow digestion. Numerous variations of this process, some including additional steps and some eliminating individual steps, are known in the art. In some embodiments, following incubation, the action of trypsin can be terminated, either by addition of a chemical inhibitor (e.g., TLCK) or by denaturation (through heat or addition of denaturants, or both) or removal of the trypsin (if the trypsin is on a solid support). There are many specific protocols available for proteolytic digestion, including automated methods using only liquid addition steps (14). In some embodiments it has been shown that automated digestion of biological samples can be very reproducible, exhibiting minimal variations (e.g., CV < 2%) between replicate samples.

[0418] In some embodiments, a desired peptide can be liberated by proteolysis without the need for disulfide reduction and alkylation (e.g., peptides that do not contain cysteine residues, and are not sterically constrained by nearby disulfide bridges), and in some cases without denaturation (e.g., peptides exposed on the surfaces of a protein). A range of alternative proteolytic enzymes can be used instead of trypsin to produce peptides defined by specific cleavage sites (including GluC, Lys-C, Arg-C, chymotrypsin, papain, pepsin, V8 protease, and the like), and chemical agents can also be used (e.g., CNBr cleavage at methionine residues).

[0419] In some embodiments a simplified digestion protocol is used comprising addition of protease to a liquid protein-containing sample without prior denaturation, reduction of disulfides or blockage of resulting cysteines. In some embodiments, heat is used to improve protein digestion by partial denaturation of protein substrates before the trypsin (or other proteolytic enzyme) is denatured, often using a variable temperature profile such as a ramp from room temperature to a higher temperature (e.g., 70C for a plasma sample). Such protocols are not expected to result in complete digestion of most proteins, but they can reproducibly generate certain tryptic peptides from some regions (e.g., surface exposed sequence segments) of some proteins, and if these peptides satisfy the requirements of TARGET peptides in a given application, the abbreviated protocol allows substantial simplification of the sample preparation workflow.

[0420] In some embodiments digestion can be carried out by immobilized proteolytic enzymes such as trypsin. Trypsin has been immobilized in the art at very high concentrations (e.g., on derivatized porous nylon, PVDF or nitrocellulose membranes) and used to perform very rapid (e.g., < 1 minute) digestion of proteins.

[0421] Proteolytic digestion disrupts protein: protein interactions by largely if not completely eliminating tertiary structure when a large protein is reduced to short peptides free to diffuse apart. This conversion of a large complex protein molecule to a series of short peptides offers a significant improvement in protein quantitation, since it removes the primary sources of assay interferences observed with immunoassays (in which a protein: protein interaction that blocks an epitope used by an assay antibody can result in a false negative result, while false positives can result from bridging interactions involving protein components not expected to be involved in the assay). An example in which tryptic digestion overcomes such an interference is the SISCAPA assay for thyroglobulin (55).

[0422] An average-length human protein produces about 50 peptides upon tryptic digestion, from which an assay designer can choose one or more peptides suitable for specific applications. This feature expands the range of detection alternatives compared to intact protein detection. Since intact proteins are so diverse in their physical properties, and therefore difficult to measure in many circumstances, the ability to select a proteotypic peptide from a range of alternatives as a stoichiometric surrogate for the intact protein is a major advantage of the digestion approach. It is often observed that within every “bad” protein there is at least one “good” peptide for a given application.

[0423] 7.1.1 Proteolytic production of peptides with either one or two amino groups.

[0424] In some embodiments it is advantageous for TARGET peptides to have a single n- terminal amino group (i.e., “single amino peptides”), and in this case digestion is preferably carried out using an enzymatic protocol in which most peptides of an appropriate length do not contain lysine (which contains a side chain amino group), and thus have a unique amino group at the n-terminus of the peptide chain. While approximately half of the peptides resulting from tryptic digestion have a lysine at the c-terminus, addition of Lys-N or similar enzymes that cleave n-terminal to a lysine residue can in many cases remove the c-terminal lysine from tryptic peptides, resulting in a larger proportion of “single amino” peptides of a useful length (i.e., the sum of the c-terminal arginine peptides produced by tryptic digestion and the set of lysine peptides from which the c-terminal lysines have been removed by the additional action of Lys-N). Another approach to decrease the proportion of double-amino peptides is to chemically convert the lysine epsilon-amino groups to homoarginine in a guanidination reaction with methylisourea (56). Alternatively, in embodiments in which it is preferred to have amino groups at both ends of the peptide (i.e., “double amino” peptides) digestion with Lys-C in place of trypsin typically leads to generation of peptides having both an n-terminal amino group and a c-terminal lysine with its side chain amino group.

[0425] 7.1.2 Sequential digestion steps and application to distinguish linkage sites

[0426] In some embodiments it is advantageous to link peptides to a specific type of molecule at or near the n-terminus, and link the peptide to a different type of molecule at or near the c- terminus - such an approach can be used to generate a construct in which the peptide is “in- line” between preceding and following polymeric components (such as oligonucleotides) useful in detection and identification of the construct. While a variety of chemical methods referred to elsewhere herein can be used to selectively couple molecules to the n-terminal amino group, a c-terminal carboxyl group, a c-terminal lysine side chain amino group, a fortuitously positioned cysteine sulfhydral group, etc., the limited specificity and quantitative yield of these reactions can make it difficult to quantitatively and reproducibly couple the two peptide termini to different molecular additions. In some embodiments the invention provides an improved alternative method making use of sequential proteolytic cleavages and coupling procedures comprising the following steps: 1) cleavage of sample proteins c-terminal to lysine residues (e.g., using the enzyme Lys-C) to produce “Lys peptides”, a number of which may include internal Arg residues; 2) reaction of the lysine side chain amino groups and the exposed peptide n-terminal amino groups of these peptides, in one or more steps (which may, for example, include click chemistry ligations), with a first added molecule (e.g., a first oligonucleotide); 3) removal or depletion of any remaining uncoupled amount of this first added molecule (or intermediate chemical components); 4) cleavage of the Lys peptides at internal arginine residues when present in a second proteolytic step (e.g., by addition of enzymes such as trypsin, Arg-C, etc.), thereby exposing a fresh and unreacted n-terminal amino group in the c-terminal part of those peptides that have a c-terminal lysine and such an internal arginine residue (i.e., the part of the original Lys peptide that extends from the amino acid following the arginine to the c-terminus); and 5) reaction of these fresh n-terminal amino groups, in one or more steps, with a second added molecule (e.g., a second oligonucleotide using a click chemistry linkage). These steps are illustrated in Figure 6 using a hypothetical protein sequence shown in Fig 6A containing arginine residues indicated by R, lysine residues indicated by K and other amino acids (all indicated here by X). Digestion with Lys-C produces a series of peptides shown in Fig 6B, each having an amino terminal and carboxy terminus. Reaction of the free amino groups (the n-terminal amino group and the side chain amino group of lysine residues K) with an added group Ml produces a series of modified peptides shown in Fig 6C. Subsequent digestion with trypsin produces a set of peptides shown in Fig 6D. Reaction of the n-terminal amino groups exposed by this second digestion with added group M2 results in the peptides shown in Fig 6E. Two of these peptides (indicated by boxes and large asterisks) have the different groups M2 and Ml positioned, respectively, at the n-terminus and near the c-terminus (i.e., on the side chain amino group of the c-terminal lysine), separated by a sequence of amino acids long enough to have a high probability of being proteotypic for the target protein (e.g., unique to one protein in the human proteome): these peptides represent likely choices for use in the invention for quantitative measurement of the target protein.

[0427] In the case of K (Lys-C) cleavage followed by R (trypsin) cleavage as described above, not all proteins of interest necessarily comprise an appropriate proteotypic “R.. K” sequence. Surprisingly, however, our in silico calculations on the known protein-encoding regions of the human genome indicate that 17,482 of the approximately 20,000 proteins in the human proteome contain at least one such peptide with a length of 7-31 amino acids (there being approximately 97,330 such peptides in the human proteome overall). Of these, approximately 13,470 peptides are 15-to-31 amino acids long, which is long enough to encompass multiple epitopes for BINDER recognition. In contrast to earlier peptide detection methods using mass spectrometry, in which shorter peptides are generally favored over longer peptides due to their greater MS signal strength and better separation by conventional reversed-phase chromatography, the single molecule methods used in some embodiments of the present invention are not typically biased against long peptides. As described herein, recognition of multiple epitopes in a peptide can provide increased specificity, as well as the possibility of leveraging avidity effects to improve BINDER capture efficiency.

[0428] Each of the two sequential linkages to ends of the peptide may, in some embodiments, involve a sequence of reactions, for example an initial reaction with an amine coupling reagent such as an NHS or sulfo-NHS conjugate of a click reagent (e.g., NHS-BCN, NHS-DBCO, NHS-TCO, NHS-tetrazine, NHS-azide, an NHS-alkyne, or the like), followed by reaction of the click group thus introduced with a corresponding click chemistry partner attached to the molecule to be added (e.g., reaction of a BCN-modified peptide with an azide-modified oligo tag or barcode). Other schemes for linkage of molecules to peptide amino groups are well- known in the art (e.g., direct reaction with an NHS adduct of the molecule to be added, etc.) and can be used to create the linkages described. In some embodiments, the same coupling chemistry is used in the first and second linkage steps, since the first linkage is complete and any component reagents can be removed prior to the exposure of the amino group used in the second linkage step. Alternatively, in some embodiments, the first linkage step is carried out using one or more reactive groups different from those used in the second step. For example, in some embodiments the first linkage step makes use of an NHS-BCN reagent to activate peptide amino groups which are then reacted with an azide-activated oligo to accomplish the first linkage. Following removal or depletion of the reagents involved in this first step and a second proteolytic digestion, a second linkage is carried out by activating the freshly-exposed n-terminal amino group with NHS-tetrazine and subsequent reaction with a DCO-activated oligo. Because of the general orthogonality of BNC-Azide and tetrazine-DCO click reactions, these two steps are unlikely to cross-react even if some amount of reagent persists from the first modification step. In some embodiments, the first step of the 2-step procedure is carried out to the point of activating the lysine amino group with a member of a first click pair (e.g., BCN) but without linkage to the first oligo, and a second activation step, following the second proteolytic step, is accomplished using a member of a second orthogonal click pair (e.g., tetrazine), after which both activated peptide groups can be reacted simultaneously with the respective orthogonal activated oligos (i.e, with an azide-activated oligo at the lysine site and a TCO-activated oligo at the n-terminus). In some embodiments that make use of sequential reactions (e.g., after the 2 successive proteolytic cleavages by Lys-C and trypsin as described above) to expose different peptide reactive sites (e.g., lysine and n-terminal amino groups), the overall yield of correctly modified products can be increased by removal of the modifying reagents (e.g., NHS-BCN and an azide- labeled oligo) used in the first step (e.g., coupling an oligo to the lysine amino group via BCN- azide click coupling) before executing the second cleavage to expose fresh reactive groups (e.g., n-terminal amino groups). This removal step decreases the probability that the peptide reactive groups exposed by the second cleavage (e.g., the amino terminal NH2 groups shown among the peptides of Figure 6D) will be modified in the same way as the amino groups exposed after the first digestion (e.g., addition of group Ml), and instead react only with a second reagent or reagents that introduce a different addition (labeled M2 in Figure 6E). In some embodiments, the reagents involved in adding group Ml to the peptides are removed by separation of peptides from the solution phase (e.g., by capturing the peptides on a suitable support such as a reversed phase or ion exchange support and then washing the soluble reagents away), by size exclusion separation to separate peptides from low molecular weight reagents (such as NHS-BCN, etc.), or by exposing the mixture to a solid support comprising a substantial content of free amino groups to which any un-reacted amino-modifying reagents can couple before removal of the support. A variety of magnetic beads, agarose particles and column packing materials having reactive amino groups are commercially available and can be used for this purpose.

[0429] Following such a 2-step linkage procedure, peptides whose sequences are bounded by a c-terminal lysine and an n-terminal amino acid that is immediately preceded in the protein sequence by an arginine will, with high likelihood, be modified with distinct added groups on the two termini, as desired. In the invention, specific peptides with these characteristics (i.e., a preceding arginine and c-terminal lysine) can be selected as TARGETS and efficiently incorporated in a predetermined orientation (i.e., n-term to c-term or vice versa) into constructs amenable to single molecule detection and counting. In the method shown schematically in Figures 12 and 13, an in-line construct is assembled that comprises, in order, an oligonucleotide in 5’ to 3’ orientation, a peptide in C-term to N-term orientation (opposite to the conventional method of writing a peptide sequence) and a further oligonucleotide in 5’ to 3’ orientation. When the final (3’) oligo is omitted from such a construct, the peptide’s n-terminus is exposed and available for sequential degradation by Edman (or alternative) chemistries used to read peptide sequence (e.g., Encodia or Quantum-Si technologies). Those knowledgeable in the art will recognize that the approach described allows the design and construction of hybrid molecules in which peptides, oligonucleotides and other polymers can be linked in a specified order and a specified orientation adapted to a variety of different single molecule detection technologies.

[0430] It is well-known in the art that improved proteolytic digestion can be achieved using a combination of Lys-C and trypsin (57), and that this combination can be used to advantage in sequence (58) addition of Lys-C first in a concentrated denaturant (e.g., 6M urea), followed by subsequent addition of trypsin after dilution of denaturant (e.g., dilution to 1.5M urea). However, the use of the two enzymes in sequence, with each enzyme cleavage step followed, respectively, by coupling of a different added molecule to available amino groups, is novel and provides an effective method for the assembly of oriented linear constructs. Similarly, the use of the two enzymes in sequence, with the first enzyme cleavage step followed by coupling of an added molecule to available amino groups, while leaving the n-terminal amino group created by the second enzyme cleavage step is a novel and effective method for the generation of a peptide-oligo construct having a free, unmodified n-terminal amino group available for cyclical degradative sequencing.

[0431] In some embodiments useful for nanopore detection, a linear construct is produced comprising a leading oligonucleotide, a central peptide, and a trailing oligonucleotide according to the invention. The method of the invention allows each segment to be assembled in a specific orientation as required by a detector such as a nanopore regulated by a DNA motor; e.g., a leading oligo oriented 5’-to-3’, followed by a peptide that is oriented c-terminal- to-n-terminal, followed by an oligo oriented 5’-to-3’ (as shown in Figure 12). Likewise, for detection technologies that require peptides to be immobilized via the c-terminus while retaining an unmodified amino group available for sequential degradation (e.g., by Edman chemistry), the second amino group modification can be omitted (i.e., just using sequential first digestion-modification-second digestion) to produce tethered peptides with free n-termini as required by some degradative sequencing single molecule detectors. While a c-terminal lysine is a preferred in-line linkage site in multiple embodiments of the sequential cleavage method described here, other cleavage sites can be used in place of Arg cleavage (as expected above in a second cleavage using trypsin), since many proteolytic enzymes, as well as chemical agents such as CNBr, generate a fresh n-terminal amino group when they cleave a polypeptide. A wide range of such alternative cleavage specificities are available: for example, the enzymes AspN, GluC, chymotrypsin, elastase or even relatively non-specific proteinase K can be used in combination with Lys-C or equivalent enzymes to generate different sets of double-amino peptides and extend the applicability of the approach to proteins not well-covered by the R. . K method.

[0432] A related alternative embodiment makes use of chemical linkages to peptide carboxyl groups instead of amino groups, and employs a series of steps to 1) cleave sample proteins n- terminal to Asp residues (e.g., using the enzyme Asp-N) to produce “Asp peptides” having an n-terminal Asp residue; 2) react the Asp side chain carboxyl groups and the exposed peptide c-terminal carboxyl groups of these peptides (and any internals Glu carboxyl side-chains), in one or more steps (which may, for example, include click chemistry ligations), with a first added molecule (e.g., a first oligonucleotide); 3) removal or depletion of any remaining uncoupled amount of this first added molecule (or intermediate chemical components); 4) cleavage of the Asp peptides resulting from the first cleavage at one or more selected internal residues (e.g., by addition of trypsin to cleave K or R, Lys-C to cleave Lys, Glu-C to cleave at Glu, etc.), thereby exposing a fresh and unreacted c-terminal carboxyl group in those peptides that have an n-terminal Asp and such an internal residue; and 5) reaction of these fresh c- terminal carboxyl groups, in one or more steps, with a second added molecule (e.g., a second oligonucleotide using a click chemistry linkage). In this case, it is preferred to select as TARGET peptides those that lack an internal Glu residue since this would comprise a 3rdcarboxyl group when AspN is used initially. A symmetrical situation would obtain if an enzyme with GluN specificity were used initially, and preferred TARGET peptides would be those lacking internal Asp residues.

[0433] In some embodiments, appropriate STANDARDS can be produced according to any of the 2-step methods described above by carrying out a similar set of steps using a synthetic peptide with an extended n-terminal sequence providing the same second-step cleavage site as the process applied in processing samples. In some embodiments the cognate STANDARD and TARGET constructs are distinguished by sequences incorporated into one or more oligos linked to the peptide. In some embodiments, one of more of the linkage steps in assembling STANDARD constructs makes use of a different click chemistry pair than that used in assembling TARGET constructs.

[0434] In some embodiments, multiple different 2-step peptide modification processes as described above are carried out in parallel on a sample, and their results combined to provide a collection of TARGET peptide constructs providing improved protein coverage or detection performance compared to a single 2-step procedure (e.g., the R...K method described initially).

[0435] 7.2 GENERATION OF TARGET PEPTIDE CONSTRUCTS

[0436] TARGET peptide constructs are created by linkage of TARGET tags to TARGET peptides in a sample digest. In some embodiments, TARGET tags are linked to common chemical features of peptides, for example peptide n-terminal and / or lysine epsilon amino groups. The common occurrence of such features implies that it would be advantageous to convert a large fraction, and potentially all, peptides in a digest into constructs of the form of TARGET constructs, irrespective of whether each such construct is to be measured against a STANDARD construct. The feasibility of such an approach is related to the efficiency with which TARGET tags can be coupled to a range of peptides and the cost of reagents required to modify more than a few specific peptides. Enabling the use of inexpensive, efficient reagents and methods to link TARGET tags to all digest peptides is therefore one of the objects of the invention.

[0437] 7.3 ADDITION OF STANDARD.

[0438] A TARGET peptide in a sample or set of samples is “standardized” by addition of a known quantity of its respective STANDARD (the STANDARD based on the sequence of the TARGET peptide with modification as disclosed above). The resulting “standardized sample digest” may be so standardized with respect to one TARGET peptide, or to multiple TARGETS (requiring multiple cognate STANDARDS). Multiple STANDARDS may be added together at one time, or at different times (e.g., after part of a standardized sample is analyzed, additional STANDARDS may be added to permit subsequent analysis additional TARGET peptides. In some embodiments the added quantity of a STANDARD is known in absolute quantitative terms (e.g., in grams / sample or grams / liter, in moles / sample or moles / liter, or molecules per sample or molecules per liter), and in some embodiments the amount of STANDARD added is known to be the same as, or have a defined ratio to, the amount of STANDARD added to other samples (thus allowing multiple samples to be compared on a consistent scale, particularly useful when samples run in a batch are compared with one another, or in longitudinal studies measuring changes in amounts of biomarker proteins occurring between serial samples from an individual). A sample to which STANDARDS have been added corresponding to a set of TARGET peptides is considered a “standardized sample” with respect to those TARGET peptides. In some embodiments one or more STANDARDS are added to a protein sample before or during digestion of the sample proteins to peptides. In some embodiments, one or more STANDARDS are added after digestion but prior to enrichment. In some embodiments additional STANDARDS are added to a digest sample that has been previously analyzed according to the invention for an earlier set of TARGET peptides and STANDARDS, enabling cycles of measurement for successive panels of peptides in a sample.

[0439] In some embodiments, the quantity of STANDARD added to a sample is chosen based on the amount of the respective TARGET peptide expected to be present in the sample(s). Specifically, the quantity of STANDARD may be based on the average or median amount of TARGET peptide observed or known to be present in similar samples, so that the ratio of TARGET peptide to STANDARD molecules falls in a range centered close to 1.0 (i.e., equal amounts). Depending on the variation in TARGET peptide amount in the samples, the ratio may for example range from 0.5 to 2, or 0.2 to 5, or 0.1 to 10, or 0.01 to 100. The benefit of arranging STANDARD amount based on TARGET peptide ranges is that it best avoids situations where the ratio is very large (e.g., 1,000: 1). Extensive investigations of the population ranges of clinical protein analytes (59) have shown that the observed range is protein-specific. High-abundance blood proteins such as albumin or hemoglobin usually vary by small amounts (much less than 2-fold), while acute phase proteins such as C-reactive protein (CRP) or serum amyloid A (SAA) can increase by 1000-fold in a serious infection. Since the present invention utilizes single molecule counting for purposes of quantitating the peptides (and therefore yields measurements whose precision is expected to depend on counting statistics in which precision is typically determined by the square root of the number of objects counted), a standardized sample with a 1,000: 1 ratio of TARGET peptide to STANDARD (or vice versa) would require the detector to count 1,000 times as many of the higher abundance peptide molecules as lower abundance peptide molecules (i.e., a total of 1,000 + 1 = 1001 times the minimum acceptable counts in each peptide to achieve desired precision based on established results in counting statistics and experimental data), thereby wasting counting capacity compared to a situation in which the peptides are present at closer to equal abundance (1 + 1 = 2 times the minimum acceptable counts would need to be counted). It will be clear to one skilled in the art that the optimal situation for efficient quantitation by counting TARGET and STANDARD peptide molecules is one in which the ratio is as close to 1 : 1 as practically possible. A person skilled in the art could design an assay according to the invention to measure hemoglobin using the population average value for the TARGET peptides as the STANDARD amount, while in CRP it could be preferable to set the STANDARD amount higher than the population average TARGET amount so as to better center the TARGET to STANDARD ratio for this highly inducible protein closer to 1.0.

[0440] For similar reasons, in some embodiments it is preferred that the amounts of molecules of different TARGET peptides measured together in a multiplex assay (each with its cognate STANDARD) should be as nearly equal as possible. This arrangement results in optimal precision achievable with a given total capacity for counting molecules according to counting statistics, and can be achieved by stoichiometric flattening during enrichment as described below.

[0441] In some embodiments two or more TARGET peptides are selected for a protein, yielding independent measurements of the protein’s amount that can be combined to deliver improved precision, or used to circumvent sequence (i.e., genetic) or post-translational variation in TARGET peptides in a population of samples. In general, unless a TARGET peptide sequence is repeated in a protein, different TARGET peptides from the same protein will be present in equal amounts after complete digestion (i.e., present in molar amounts equal to the molar amount of the parent protein). In some embodiments, multiple TARGETS are selected from a protein that exhibits highly variable amounts in relevant samples, and their respective cognate STANDARDS are added at different amounts so that the TARGET -to- STANDARD ratio of at least one of the TARGETS is close enough to 1 : 1 to be efficiently countable and thereby furnish an accurate ratio measurement. In some embodiments, for example, three TARGET peptides are selected and their respective STANDARDS added to the sample digest at O.lx, l.Ox, and 10. Ox the average amount of TARGET in an average sample, so that variation in the amount of target protein over a range of 10-fold above and below the expected amount (100-fold dynamic range) will be measured by at least one of the TARGET -to- ST AND ARD ratios close to 1 : 1. In some embodiments of this kind, the amounts of the respective BINDERS used to enrich the peptides are adjusted separately to bring the peptides into flat stoichiometry (near equivalence) before detection and counting.

[0442] 7.3.1 Differences between MS and single molecule requirements

[0443] It is worthwhile to note that detection of peptides by single molecule counting involves significantly different tradeoffs as compared to quantitation by mass spectrometry (MS). MS typically produces analog measurements of the amount of a molecule that passes through a set of mass filters, or reaches a detector after some mass or size-based separation process. Since different molecules are typically detected at different times, the amount of one peptide does not usually affect the detection of a different peptide in a significant way, apart from extreme cases in which a detector is saturated, or a total aggregate ion capacity is exceeded. The dynamic range of modem triple-quadrupole MS instruments approaches 100,000-fold for a given analyte (e.g., peptide), and the MS (or LC-MS) system can accommodate two different peptide molecules that differ in abundance by 1 million-fold and still produce some quantitative information on both (e.g., by selecting a low-efficiency detection mode such as an infrequent MRM fragment mass for the high abundance molecule and a very efficient detection mode for the low abundance molecule). The detection process is typically driven by a chromatographic separation of 2 to 60 minutes prior to introduction of separated peptides into the MS, and is essentially insensitive to the total number of molecules in the applied sample: the analytical run will “consume” all the applied sample and will occupy the same period of time regardless of the number of molecules being analyzed. The situation pertaining in single molecule detection and counting is very different: the number of molecules sequenced and counted depends directly on time. Thus, for nanopores the number of molecules analyzed is a direct function of the time required to observe a typical molecule’s sequence (e.g., 0.25sec for a 50bp equivalent length oligo:peptide construct at a typical rate of 200bp / sec through a single nanopore) multiplied by the number of pores operating in parallel: running the device twice as long will typically detect twice as many peptides. Hence nanopore detection methods are inherently limited by the number of molecules that can be sequenced per time. As discussed below, the precision of ratios between TARGET and STANDARD peptides is largely determined by counting statistics, with more counts yielding higher precision. As a result, it is highly desirable in the present invention to avoid wasting capacity (in peptides or time) in counting more than a specified number of a given peptide, since this capacity could instead be used count more molecules of a lower abundance peptide to improve its precision. In some embodiments this principle is applied to generate enriched peptide samples in which a) the number of added STANDARD molecules is close to the average number of TARGET peptide molecules in typical samples, and b) the sum of TARGET + STANDARD molecules is approximately the same for all proteins and peptides being measured in a multiplex panel. This principle, referred to herein as “Stoichiometric Flattening” is described in greater detail below.

[0444] Similarly, for degradative peptide sequencing methods, a series of sequential steps is applied in parallel to a large number of immobilized peptide molecules, with the number of molecules being determined by the physical scale of the device used (number of peptides that can be immobilized and resolved by the detector) and the number of such runs. Likewise, for single molecule imaging detection methods, a large but limited number of immobilized peptide molecules is detected in a run, with the number of molecules being determined by the physical scale of the device used (number of peptides that can be immobilized and resolved by the detector) and the number of such runs. The overall throughput in terms of the number of molecules analyzed per unit of time is determined by the number of molecules sequenced per run, the duration of a run, and the number of runs (for a batch method). Hence, as with nanopore sequencing, efficiency is maximized by sequencing similar numbers of each TARGET and STANDARD peptide, instead of allowing one or more high abundance peptides to occupy a large fraction of the capacity, reducing the numbers of molecules from lower abundance TARGETS and thus the precision of their measurement.

[0445] 7.3.2 Comparison between samples

[0446] In some embodiments, a fixed amount of each STANDARD (e.g., an equal volume aliquot from the same STANDARD stock solution) is added to each of a multiplicity of samples, thereby establishing a shared reference basis that allows accurate comparison of the amounts of TARGET peptides between these samples. This approach enables relative comparison of protein amounts between samples, but does not directly provide absolute quantitative information (e.g., in mass or concentration units) without the use of external calibrators (see e.g., US provisional patent application 63 / 213,371 - entitled Calibration of Analytical Results in Dried Blood Samples, filed 6 / 22 / 21, incorporated herein in its entirety). In some embodiments, the amounts of STANDARD molecules added represent known quantitative amounts (i.e., numbers of molecules, mass, or concentration), in which case the absolute amounts of TARGET peptides can be estimated by multiplying the STANDARD amounts by the observed TARGET peptide to STANDARD ratios.

[0447] In some embodiments, STANDARDS are generated from synthetic (e.g., recombinantly expressed) protein constructs whose digestion yields STANDARDS in relative ratios defined by their copy number in the construct’s sequence (see for example Patent Application US 2006 / 0154318). In some embodiments the STANDARDS are provided in physical forms enabling simplified manipulation and addition (as described in US9274124). In some embodiments the STANDARDS are added as peptides in solution.

[0448] 7.3.3 Amount of standard added

[0449] In some embodiments, the amounts of each STANDARD added can be determined according to (or equal to) the average or baseline levels of corresponding TARGET peptides observed in a subject’s prior samples, thus providing STANDARD levels that are more or less equal to the expected TARGET peptide levels for that subject. This approach provides optimal precision and efficiency for the measurement of longitudinal samples from that subject, and represents an ideal case of personalized protein measurement with the potential to maximize precise detection of small protein changes from baseline levels over time. 7.4 ENRICHMENT OF TARGET PEPTIDES AND STANDARDS.

[0450] Once a sample digest has been “standardized” by addition of one or more of the appropriate STANDARDS in the desired amounts, the sample can be fractionated or purified to enrich analytes to a desired level, and / or to deplete unwanted components (the analytical “matrix” background). In some embodiments, peptide-oligo constructs according to the invention are purified by a process that reversibly captures peptides (e.g., reversed-phase adsorbents such as Cl 8 resins) with the result that oligonucleotides and other non-peptide components not part of peptide-oligo constructs can be washed away and thus removed. In some embodiments, peptide-oligo constructs according to the invention are purified by a process that reversibly captures oligonucleotides (e.g., adsorbents such as Ampure resins) with the result that peptides, remaining proteins and other non-peptide components not part of peptide-oligo constructs can be washed away and thus removed. In some embodiments addition features of the peptide-oligo constructs are used to isolate them, for example binding of a biotin group engineered into the oligo, the peptide or the linkage between them can be captured by immobilized streptavidin as a means of cleaning up the desired constructs.

[0451] After any such construct cleanup steps, enrichment using specific affinity BINDERS is an important aspect of the invention. In some embodiments BINDERS are used to carry out specific affinity enrichment of the respective cognate TARGET peptide and STANDARD pairs. The TARGET peptide, its STANDARD and the BINDER designed to bind them collectively form a cognate set of molecules specialized for the measurement of a specific TARGET peptide and thus its parent protein.

[0452] 7.4.1 Magnetic bead enrichment

[0453] In some embodiments, BINDERS are immobilized on magnetic beads and these beads mixed with the standardized digest and incubated to allow binding of peptides to the BINDER. In some embodiments (e.g., using anti-peptide antibodies as BINDERs) the BINDERS can be bound to commercially-available Dynabeads G via Protein G’s affinity for the Fc domain of IgG, and optionally covalently linked to the beads using DMP crosslinking or other equivalent crosslinking methods forming bonds between the magnetic bead and the BINDER. In some embodiments BINDERS are bound to other types of magnetic particles such as Tosyl-activated beads, or otherwise chemically bound to particles that can be manipulated to allow exposure to and removal from a sample. Due to their specific affinity for the TARGET peptide sequences, the BINDERS bind the respective TARGET peptides and STANDARDS when placed in contact with them. In some embodiments, BINDERS are in contact with a standardized sample digest for a 30-minute incubation period with shaking to keep the beads suspended. In some embodiments, BINDERS are incubated with standardized digest for shorter periods (e.g., 1, 2, 5, 10, 15, 20 or 25 minutes) and in some embodiments, BINDER are incubated with standardized digest for longer periods (e.g., 45, 60, 90, 120, or 180 minutes, or 4, 5, 6, 9, 12, 18 or 24 hours). Depending on the kinetic properties of specific BINDERS, the abundances of their cognate TARGET and STANDARD peptides, and the presence or amounts of any competing sample peptides (e.g., peptides with different but similar sequences to a TARGET peptide), persons skilled in the art will understand how to perform experiments to evaluate and select a suitable incubation time. After binding, the beads with attached BINDERS and their bound peptide cargo are separated from the digest. To achieve this separation, the BINDER beads can be collected together using a magnet and either removed from the digest (for example using a Kingfisher device provided by ThermoFisher); held in a vessel (for example by magnetic attraction to the side of a well of a 96-well plate) while the digest solution is removed to another container by a pipetting device (e.g., an Agilent Bravo, Beckman Counter Biomek, Hamilton, Tecan or other liquid handling robot); held magnetically in a conventional pipette tip while surrounding liquid is expelled and replaced (e.g., as in the established “Magtration” technology); processed in a “magnetic bead trap” (60) or processed manually (e.g., by manipulation of vessels, magnets and handheld pipettes). The standardized digest remaining after separation from BINDERS may be preserved apart from the BINDERS and stored or subjected to additional processes to measure additional constituents at a later time.

[0454] In some embodiments, the beads with BINDERS and specifically bound peptide cargo are washed by addition to, and mixing with, aliquots of a wash solution, after which the beads are recollected and separated from the wash. In some embodiments the beads are washed 1, 2, 3, or 4 times with volumes of 50 to 400uL of wash solution, which may include buffers (e.g., PBS or Tris, typically at pH between 6.0 and 8.5 when antibody BINDERS are used), gentle detergents (e.g., CHAPS or deoxycholate), and / or low concentrations of an organic solvent (e.g., 5-20% acetonitrile added to help remove peptides bound to beads non-specifically). Persons skilled in the art will be able to evaluate and select wash solution compositions that are most effective at removing non-target digest peptides (and other components) while retaining TARGET peptides and STANDARDS on the BINDER. Since it is desired that the specifically-bound peptides remain attached to the beads during the wash procedure, in some embodiments the BINDERS are designed or selected to have half-off-times (the time period over which half the bound molecules become unbound, i.e., the dissociation half-life) longer than the time required to execute the sequence wash steps (typically 10-15 minutes using current laboratory automation systems). In some embodiments, for example those in which the BINDERS are present at high local concentrations at some point(s) during the wash process, TARGET and STANDARD peptides that leak from the BINDER can be re-bound by the same or other BINDER sites before being lost. It will be understood by those skilled in the art that experimentation with specific BINDERS, TARGETS and STANDARDS, specific wash solution compositions and temperatures, specific wash volumes and vessel geometries, and specific sample digest matrices is required to optimize a) the enrichment of the TARGETS and STANDARDS and b) the removal of the other digest components. As a general matter, the purer the TARGETS and STANDARDS are after enrichment, the better the invention will function to measure them precisely.

[0455] In some embodiments, the recovery of TARGETS and STANDARDS from a digest by enrichment using BINDERS is evaluated by successively contacting 2 or more separate aliquots of BINDER with the digest, and comparing the amounts eluted from the first and second BINDER aliquots. In cases where the number of BINDER binding sites is greater than the number of TARGET or STANDARD molecules, an effective BINDER will typically capture 80% or more of these peptides on the first capture step, and the second (and possible subsequent) capture steps will capture successively smaller amounts of the peptides. The ratio of TARGET and STANDARD captured in the first capture divided by the amount of these peptides in the sum of first and second captures provides a useful index of the overall recovery. In cases where the number of BINDER binding sites is less than the number of TARGET and STANDARD molecules, sequential BINDER captures will yield more nearly equal amounts of these peptides. In some embodiments, BINDERS immobilized on magnetic beads are contained within microfluidic systems capable of moving the beads between different liquid volumes to effect the steps of the invention. Microfluidic systems and technologies well known in the art allow use of reduced liquid volumes (and thereby less dilution of low-concentration analytes such as enriched peptides) and more complex, multi-step chemical processes with reduced losses as compared to conventional lab-scale (i.e., 5-500uL) liquid handling processes (for example in a magnetic bead trap device (60) .

[0456] 7.4.2 Column format enrichment

[0457] In some embodiments, BINDERS are immobilized on columns through which sample digest and wash solutions can be passed (3), typically using a liquid chromatography system.

[0458] 7.4.3 Post-peptide-capture immobilization

[0459] In some embodiments BINDERS are contacted with a standardized digest in solution (i.e., BINDERS free in solution, not bound to a support), thereby maximizing freedom of diffusion and potentially providing faster binding to TARGET and STANDARD peptides or respective VEHICLE constructs. After binding, the BINDERS can be themselves captured on magnetic beads (e.g., protein G coated beads in the case of antibody BINDERS, streptavidin coated beads in the case of BINDERS that have been previously biotinylated, etc.) or on columns functionalized with equivalent capture functionalities.

[0460] 7.4.4 Multiple enrichment cycles

[0461] In some embodiments, peptides captured by and eluted from BINDERS are subjected to one or more additional cycles of BINDER enrichment. When bound peptides are eluted from BINDERS by a change to specific elution solution conditions (e.g., pH 2.5 in the case of antibody BINDERs), reversal of these conditions (e.g., neutralization to pH near 7.0) can allow the peptides to bind to respective BINDERS once again. Similarly, if elution is carried out by exposure to a chaotropic salt, a detergent, or increased temperature, these conditions can be reversed (e.g., through dilution of a salt or detergent, dialysis, or cooling), restoring conditions in which binding to BINDER (typically fresh BINDER) can occur. After the initial BINDER enrichment process, the captured peptides are removed from the bulk sample digest, and thus the vast majority of non-target “matrix” peptides are no longer present. A second BINDER enrichment cycle thus begins with a much smaller amount of total peptide material, in which the TARGET and STANDARD peptides represent a much larger fraction of dissolved material (mainly peptides). In some embodiments, a second BINDER enrichment cycle is carried out using fresh BINDER (i.e., BINDER that has not previously been exposed to complex digest), and this additional cycle further depletes non-target matrix peptides, while recovering a large fraction of the TARGET and STANDARD peptides, and resulting in a purer sample of the TARGET and STANDARD peptides of interest. Increasing the fraction of peptides that are desired TARGET or STANDARD peptides directly improves the efficiency of single molecule detection aimed at measuring those peptides by decreasing the time and resources spent sequencing other peptides. In some embodiments, one or more additional enrichment cycles are carried out using a smaller amount of BINDER (e.g., a smaller volume of beads) than used in the initial enrichment cycle, resulting in an opportunity to reduce the volume in which peptide constructs are eluted and thereby increasing the concentration of TARGET and STANDARD peptide constructs introduced into a detector. The ability to deliver TARGET and STANDARD peptides in a small volume, or on a small number of beads, can improve the efficiency of single molecule detection. Providing peptides for single molecule detection in as concentrated a form as possible can be important in specific detection methods, for example in delivering peptides to the vicinity of a sequencing nanopore as described below. In some embodiments, a first BINDER capture step is carried out using BINDER immobilized on a large number of small (e.g., 1, or 2.8 or 5 micron diameter) magnetic beads, thus maximizing the dispersion of BINDER in the sample digest volume and decreasing the diffusion distance and time required for peptide capture, after which the captured peptides are eluted and recaptured by fresh BINDER immobilized on a smaller number of beads, with the result that the peptides captured in the second round are both purer (fewer non-TARGET peptide molecules) and more concentrated (e.g., when the beads are collected magnetically into a mass for removal from contact with the peptide-containing liquid). In some embodiments, recovery of TARGET and STANDARD peptides in a very small volume of beads allows the bound peptides to be exposed to very small volumes (equal to or slightly greater than the included volume of the bead mass) of reagents used in the preparation of peptides for linkage to VEHICLES as described below. In some embodiments a second capture makes use of BINDER bound to a small number of larger (e.g., 10-40 micron diameter) beads each having a greater BINDER capacity, such that each bead can be taken through the series of chemical steps to prepare it for and / or complete VEHICLE linkage to it in a separate container, as described below.

[0462] In some embodiments a first BINDER enrichment cycle is carried out to recover and purify the TARGET and STANDARD peptides from a mass of sample digest, and the same or similar peptide-specific binders used to identify these TARGET and STANDARD peptides in a single molecule detection system. The first BINDER enrichment cycle can thus serve to remove non-TARGET peptides present in the digest (thus minimizing analytical capacity wasted on irrelevant peptides), and optionally to improve the stoichiometric flatness of a series of different TARGETS to be detected and counted (as described in detail below). For example, a fluorescently-labeled BINDER can be used to detect its cognate TARGET and STANDARD peptides in an optical imaging system of the kind used in high-throughput DNA sequencing or in similar protein detection systems (e.g., US 2021 / 02397, and thereby used to count the numbers of such molecules. In some embodiments a second class of BINDER that specifically recognizes a unique tag present in the STANDARD but not in the TARGET peptide is used to separately distinguish the STANDARD peptide molecules from the TARGET molecules. Using these two separate detection steps applied to a population of single molecules (i.e., identification of the molecules in a TARGET + STANDARD pair, and separate identification of STANDARD peptides), allows separate identification and counting of the molecules of TARGET and STANDARD peptides, and thereby determination of the TARGET-to- STANDARD ratio. In some embodiments, this approach is used to identify all STANDARD molecules in one recognition step (using a BINDER specific to the STANDARD molecule tag), and each different TARGET+ STANDARD pair is identified by a separate detection step using the TARGET-specific cognate BINDER. In some embodiments, the tag distinguishing the STANDARD molecules can be an added amino acid sequence on either amino or carboxy terminus of the TARGET peptide sequence (for example the well-known FLAG peptide sequence used in recovering expressed proteins), a chemical group bound to the STANDARD (such as biotin), or any of a variety of distinctive chemical structures unlikely to be found in the group of TARGET peptides. In some embodiments this approach can be applied to count whole protein molecules instead of proteolytic peptides using peptide-sequence-specific BINDERS to identify their cognate linear sequence epitopes in intact target protein molecules. In such embodiments aimed at detecting protein molecules, a first BINDER enrichment cycle can use BINDERS specific for an intact protein as well as BINDERS specific for linear peptide epitopes, and STANDARDS can be versions of the intact protein with any of a variety of unique tags as for peptides.

[0463] 7.4.5 Stoichiometric flattening

[0464] In some embodiments, a plurality of TARGET and STANDARD peptides is enriched by the corresponding plurality of BINDERS, and the relative amounts, kinetic properties or solution conditions of the BINDER enrichment are selected or adjusted so as to accomplish some degree of stoichiometric flattening; i.e., to diminish differences in the relative amounts of different TARGET + STANDARD peptide pairs. As described above, to obtain the benefit of stoichiometric flattening, it is necessary to standardize measurement of TARGET peptide constructs by incorporating STANDARD constructs that can act as internal standards before BINDER enrichment (to preserve information on quantitation), and to effect enrichment using peptide-specific BINDERS.

[0465] Stoichiometric flattening, described in greater detail below, is distinct from conventional sample preparation enrichment methods. In genome sequencing, gene stoichiometries are equal, or close to equal, to begin with and do not need to be adjusted. In studies of RNA’s, which can be present at a range of relative abundances, enrichment can be used to focus on a specific set of RNA’s (e.g., mRNA’s), but sequence-specific internal standards are not used and thus differential enrichment of different sequences destroys relative abundance information. In proteomics studies aimed at broad proteome coverage, internal standards have found more use because of the practicality of using stable isotope labeled peptides and mass spectrometric detection; however, the requirement for individual enrichment BINDERS for a significant fraction of digest peptides remains an insurmountable problem given the number of proteins present in most samples (the ~50-fold higher number of peptides therefore present in a sample digest). In the case of targeted methods such as SISCAP A, the use of mass spectrometric detection makes it possible to achieve some degree of stoichiometric flattening by choosing target peptides that have very different ionization properties; i.e., choosing peptides with extremely high MS detection performance to represent low abundance proteins (where sensitivity is a major challenge) and choosing peptides with much lower performance to represent higher abundance molecules. Such peptide choices are an important component of stoichiometric flattening in MS methods, in addition to the adjustment of relative BINDER enrichments as described above.

[0466] Stoichiometric flattening by choosing peptides with different detection efficiencies is not useful in single molecule counting methods, which generally count all molecules with equal efficiency. In the present invention, focused on single molecule detection and counting, in most embodiments there is little or no difference expected between the detection efficiencies of different peptides (in contrast to the situation in mass spectrometry where ionization efficiency, ion transport, molecular fragmentation, and detection efficiency are all peptide- specific and highly variable). While the near-equivalence of peptides as far as detection efficiency in single molecule methods is an important advantage in expanding peptide choice for any given protein target, it removes one of the major avenues available for stoichiometric flattening and restricts the method to adjustments in the BINDER capture step only.

[0467] 7.4.6 Consequences of not flattening stoichiometry

[0468] A further important distinction between the methods based on mass spectrometry versus single molecule sequence-sensitive detectors is the impact of failure to flatten stoichiometry. In typical mass spectrometry protocols using liquid chromatography (LC-MS), a sample is analyzed using a method involving a specified chromatographic separation (with a specified duration, usually in the range of 1-60 minutes), and the mass spectrometer is presented with whatever peptide ions emerge from the end of the column, whether they are too few to register as a signal, or too many to be accurately measured, or in between. In other words, the sample analysis typically takes a pre-specified length of time whether the stoichiometry of TARGET peptides in the sample has been flattened or not. Using single molecule methods, for example using nanopore sequencing, each peptide molecule takes some time to be analyzed - during this time one pore is occupied with one molecule, or a fixed number of molecules requires a given block of time to analyze. While the throughput of such a process can be increased by providing multiple pores or molecule immobilization sites, or decreasing read time, there remains a direct relationship between the number of molecules detected and the time devoted to the analysis of that sample. It is therefore of paramount importance when using single molecule methods for quantitation according to the present invention that the system not spend unnecessary time sequencing a) peptide molecules that are not TARGET or STANDARD peptides (i.e., do not contribute to the results desired), or b) peptides that have already been sequenced in sufficient numbers to provide a TARGET-to- STANDARD ratio with the desired precision (e.g., %CV based on counting statistics). Removing peptides defined by (a) above is a matter of enriching TARGET peptides and depletion of the rest. However, minimizing the counting of peptides that are surplus to statistical requirements (those defined by (b) above) is a key benefit of stoichiometric flattening. It is thus key to the practicality of single molecule detection methods for quantitation of selected TARGET peptide targets, and can reduce the total number of molecules that need to be counted in practical biomarker studies by large factors (e.g., 16,000-fold improvements in efficiency in measuring components of dried blood spots, as described in more detail below - “Stoichiometric Flattening”).

[0469] 7.5 PREPARATION OF PEPTIDES FOR SINGLE MOLECULE SEQUENCE- SPECIFIC DETECTION.

[0470] In some embodiments the detection of BINDER-enriched TARGET and STANDARD peptides can be facilitated by certain chemical modifications, including covalent linkage to polymeric molecules on one or both ends (i.e., on or near peptide n- or c-termini), or linkage to a support, surface or bead, resulting in constructs with improved uptake and sequence readout by a sequence-sensitive detector, and / or incorporating additional information beyond the peptide itself in the form of detectable polymer sequences (e.g., DNA sequence tags).

[0471] 7.5.1 Chemical modification of peptides on BINDER

[0472] In some embodiments, peptides are chemically modified while bound non-covalently to a BINDER (e.g., a BINDER that is used to enrich them from the peptide sample). Modification of the peptides while thus non-covalently anchored to a BINDER (which may itself be bound to a solid support) facilitates exchange of reagents between steps of a multi- step series of chemical modifications, avoids the necessity for other more cumbersome purification methods between steps (e.g., to separate modified peptides from reagents and unmodified peptides), and allows the peptides to be concentrated when necessary (e.g., by gathering magnetic beads bearing the BINDER into a solid mass with minimal included liquid). These advantages are very useful when a single modification step is required, and progressively more valuable as more modification steps are needed.

[0473] In some embodiments a series of chemical steps are used to prepare peptides for analysis, and it may be necessary to remove the chemical reagents required for one step, and in some cases wash them way, before adding reagents for the next step. When the off-rate of the BINDER is low, with a half-off-time for example longer than 10-15 minutes (as is typical for antibody BINDERS developed for use in SISCAP A), one or more rapid chemical reactions can be carried out before peptide molecules dissociate from the BINDER. In some embodiments involving more time-consuming chemical modification processes (e.g., requiring an incubation period of 1-60 minutes for a reaction to progress towards completion), BINDERS can be concentrated (e.g., by collecting magnetic beads into a mass or small volume, or using a column format bearing a high density of immobilized BINDERs) during steps of the process, and during these periods any peptide that dissociates from a BINDER is likely to quickly rebind to another BINDER site given the high local BINDER concentration. This kinetic effect effectively prolongs the time available for chemical modification of BINDER- bound peptides.

[0474] Certain limitations apply to this approach in some embodiments. For example, if the BINDER is an antibody, then solution conditions that would denature the antibody (e.g., strong detergents such as SDS at high concentrations, extremes of pH, or high temperatures) or cause the peptide to be released from the antibody (e.g., pH below 3.5 or presence of 2M NH4SCN) can be problematic while carrying out the desired peptide modification. However, means are known in the art for carrying out a wide range of desirable modifications to TARGET and STANDARD peptides under solution conditions compatible with retention of the peptides on a cognate BINDER.

[0475] In some embodiments the BINDERS themselves are modified prior to use in capturing TARGET and STANDARD peptides in order to prevent or diminish their reaction with reagents intended to react with the peptide cargo. Thus, in some embodiments, some or all of any free amino groups on an antibody BINDER can be blocked, for example by PEGylation using commercial NHS-PEG reagents. In some embodiments using DNA or RNA aptamers as BINDERS, modifications may be required to prevent hybridization between the aptamers and sequences being attached or ligated to peptides on a BINDER.

[0476] In some embodiments, the peptide modifications can be carried out while the peptides are bound to a support by a general but less-specific mechanism (e.g., to a reversed-phase support such as C18 particles), or free in solution.

[0477] 7.5.2 Amino group linkage

[0478] Among the most useful reaction sites on a peptide is an amino group. For peptides generated by digestion with trypsin (the most commonly used proteolytic enzyme), all correctly processed peptides (except a protein’s c-terminal peptide) should have a c-terminal lysine or arginine residue. Since lysine is the only amino acid with a side-chain primary amino group (reaction with which would allow two attachment sites on a peptide), some embodiments make use of TARGET peptides from the group of tryptic peptides with c-terminal lysine when preparing a construct with polymer additions at both ends (a “Double amino” peptide), or selecting TARGET peptides from the group of tryptic peptides with c-terminal arginine (which peptides will have a single n-terminal amino group available to react with a linker; i.e., a “Single amino” peptide) when an extension on only one end (the n-terminus) of the peptide is desired.

[0479] In some embodiments, an advantage of linkage through amino groups, particularly in the case of c-terminal lysine peptides modified by a linkage chemistry that results in a decrease in the peptide’s net positive charge, is that the peptide subsequently has little or no positive charge (e.g., if it contains no His resides), and at least one (the c-terminal carboxyl) and perhaps more negative charges (if the peptide contains Asp or Glu amino acids). Peptides with a net negative charge have the same charge polarity as nucleic acids (negative, on account of the phosphate groups), facilitating the movement of both types of polymers through a pore using the same polarity electric field.

[0480] 7.5.3 Carboxyl group linkage

[0481] In some embodiments linkage through peptide carboxyl groups can be used (61) but this approach has limited ability to distinguish between c-terminal carboxyls and side chain carboxyls of aspartic and glutamic acids, and thus could present additional constraints on peptide selection (i.e., de-selection of Asp and Glu containing TARGET peptides) or give rise to multiple constructs due to side reactions. In some embodiments, TARGET peptides devoid of aspartic and glutamic residues, and hence having a unique carboxyl group at the peptide c- terminus, are used with carboxyl coupling chemistries well-known in the art to link peptides through the c-terminus.

[0482] 7.5.4 Other linkage sites

[0483] Other linkage sites and reaction chemistries can be employed. Linkage through a cysteine sulfhydryl group is frequently favored when a peptide’s sequence can be freely designed - however the occurrence of a cysteine residue at the n-terminus or c-terminus of a proteolytic peptide in a natural protein is infrequent, thus representing a limiting constraint on TARGET peptide selection.

[0484] Chemistries are known in the art for specific chemical modification of, and / or linkage to, histidine, tyrosine, tryptophan and other amino acids, and these can also be used in the invention.

[0485] 7.5.5 Peptide:Oligo in-line constructs using sequential amino linkages

[0486] Figure 14 illustrates an embodiment for modification of enriched TARGET and STANDARD molecules (either of these being labeled a “Peptide” in the Figure) by linkage of an n-terminal amino group to a single-stranded oligonucleotide Leader (labeled Oligo 1) and an epsilon-amino group of a c-terminal lysine to a single-stranded oligonucleotide Trailer (labeled Oligo 2). Such a hybrid molecule can be considered an “in-line” peptide- oligonucleotide construct; i.e., one in which the construct forms a continuous backbone with “side-groups” consisting primarily of bases and amino acid side chains. In the embodiment shown in Figure 14, the linkages are carried out in a series of steps using “click” chemistry, a bio-orthogonal reaction chemistry (25, 62, 63). First, (Step 1 in Figure 14) a peptide’s n- terminal amino group is reacted with azidoacetic anhydride (AAA) under conditions of pH (e.g., pH 5.5 to 7.0, preferably 6.7) that effectively prevent or reduce reaction with a lysine’s tertiary amino group (25, 64), resulting in the introduction of an azide functionality at the peptide’s n-terminus (shown in Step 2 of Figure 14). After removal of the AAA (e.g., by washing the beads carrying the BINDER and their peptide cargo), this azide group is then allowed to react with an amount of oligonucleotide Oligo 1 (Step 3 in Figure 14), which has been prepared with an aza-dibenzocyclooctyne (ADIBO) functionality on the 3’ end, forming a “click” chemistry linkage of the Oligo 1 to the peptide n-terminus (25) (Step 4 in Figure 14). In a next step, after removal of unreacted Oligo 1, the peptide’s c-terminal lysine tertiary amino group is allowed to react with an azidoacetic acid NHS ester (Step 5 in Figure 14), thereby introducing an azide functionality on the peptide’s C-terminal lysine side chain (Step 6 in Figure 15). After removal of the unreacted NHS ester (e.g., by washing the beads carrying the BINDER and their peptide cargo), this azide group is then allowed to react with an amount of oligonucleotide Oligo 2 (Step 7 in Figure 14), which has been prepared with an aza- dibenzocyclooctyne (ADIBO) functionality on its 5’ end, forming a “click” chemistry linkage of the Oligo 2 to the peptide’s C-terminal amino acid (Step 8 in Figure 14). The result is a peptide-oligonucleotide construct comprising a peptide linked to an oligonucleotide on each end. Carrying out the additions step-wise as shown, using the two Oligos, each prepared with a “click” linkage site on one end, allows the construct to be generated with specified oligo polarities (i.e., 5’ to 3’ or vice versa) in each case, as may be necessary to achieve the desired interaction with DNA motors, pores, etc. The Oligos can each be single-stranded, or they can be rendered duplexes by hybridization with complementary sequences over part or all of their length(s).

[0487] In some embodiments the locations of the “click” reactive pairs can be inverted (e.g., modifying one or both of the peptide ends with ADIBO, and the Oligos with azide). Linkage at the peptide c-terminus can alternatively be introduced through modification of the peptide c-terminal carboxyl group (e.g., a c-terminal arginine, in which case the n-terminal amino group is the only amino group in the peptide) instead of a lysine tertiary amino group.

[0488] In some embodiments some or all of the steps of the double ligation scheme shown in Figure 15 are carried out with peptides bound to a BINDER on magnetic beads. Using peptides longer than the groove of a typical BINDER binding site (typically 4-8 amino acids), and BINDER that binds an epitope that does not include either peptide terminus, leaves both peptide termini available for reaction as shown. Magnetic beads carrying the BINDER and bound peptides can be exposed stepwise to the sequence of reagents as shown by moving the beads from one reagent solution to the next, with optional wash steps in between as required, or alternatively the beads can be gathered to the side of a vessel using a magnet and the sequence of reagent solutions added and removed, or manipulated in a microfluidic device.

[0489] Many alternative chemical reaction schemes are known in the art that can be used in place of the specific reagents and steps shown in Figure 14, while generating a construct with similarly useful features. A variety of reagents exist that are capable of introducing “click” chemistry functional groups into peptides and oligonucleotides (17). Examples include azide and tetrazine functionalities that are capable of specific, bio-orthogonal reactions with alkyne functionalities, some requiring Cu(I) catalysis (which is less preferred in some embodiments), or strain-promoted alkyne cycloaddition (SPAAC) reactions with bicyclononyne (BCN) or cyclooctyne and derivatives such as dibenzocyclooctyne (DIBO) or Aza-dibenzocyclooctyne (ADIBO shown in Figure 14).

[0490] In some embodiments one or both of the oligos linked to the peptide is part of a multimolecular construct designed to facilitate nanopore sequencing. Figure 15 illustrates a modification of the embodiment of Figure 14 in which the oligo joined to the N-terminal amino group of a peptide (Oligo 1) is part of a duplex having a site at which a “motor protein” can be (or in some embodiments already is) bound. The overall construct shown in Figure 15 provides a leading oligo motor construct capable of controlling movement of the peptide through the pore (providing a stepwise ratchet motion as used, for example, in the commercially-available Oxford Nanopore sequencing platform “Y-adapter”; https: / / nanoporetech.com / sites / default / files / s3 / literature / product-brochure.pdf), followed by a peptide to be sequenced, and a trailing oligo 2. A tether molecule that hybridizes to some sequence in the construct may be used to associate the construct with the membrane in which the nanopore is located, thereby increasing the Leader’s probability of entering the nanopore.

[0491] In some embodiments, alternative chemistries can be used to join the peptide and oligos, such as use of an NHS-functionalized oligo that can react directly with a lysine or n- terminal amino group.

[0492] Similar peptide-oligonucleotide constructs have been described, for example in WO 2021 / 111125, using different “click” reagent combinations and using alternative means of purifying desired products (e.g., purifying via the oligo part of the construct using Agencourt AMPure XP beads). The referenced disclosure showed that oligo-peptide-oligo constructs assembled by click chemistry can be processed by existing nanopores and DNA motors to generate reproducible ion current traces indicative of peptide sequence. However, these disclosed inventions do not include the use of BINDERS to carry and / or move TARGET peptides during one or more steps of oligo-peptide construct assembly, washing and purification, and do not include STANDARDS or use of BINDERS for enrichment or stoichiometric flattening.

[0493] 7.5.6 2-step digestion to distinguish amino groups

[0494] As described above and illustrated in Figure 6, a sequential enzymatic approach is used in some embodiments to derivatize peptides with two different added molecules, resulting in an ordered construct. Using this approach, a peptide can be linked to a specific oligo on the n- terminus and a different oligo on the c-terminal lysine reside. This approach is not completely general since it requires a specific order relationship between an arginine immediately preceding the peptide sequence and the peptide’s c-terminal lysine. Such peptides are common but not universal in proteins, and when added to a requirement such as uniqueness in the human proteome, it is probable that only a subset of proteins can be represented by proteotypic peptides having such a structure.

[0495] 7.5.7 Delivery of peptides to a site of detection.

[0496] The potential of single molecule detection to improve assay sensitivity is enormous provided that the desired peptide molecules can be efficiently presented to the sequencing machinery - i.e, to occupy a large fraction of the available peptide sequencing capacity. As is well known in peptide analysis by mass spectrometry, peptides at very low concentration (i.e., at low single molecule levels) easily become “lost” through escape into dilute solutions, binding to surfaces, etc., and therefore detection can become more challenging as target abundance or concentration is reduced.

[0497] Some embodiments therefore make use of the localization of TARGET and STANDARD peptide constructs on BINDERS, which themselves can be bound to particles such as magnetic beads, to transport the peptides (e.g., as peptides, or as part of peptide-oligo constructs as disclosed above) into close proximity to the site of detection (e.g, a nanopore, a surface on which they may be imaged, or a well in which they may be subjected to degradative sequencing). Recent progress in this area is disclosed in patent applications US 2020 / 0284783 and US 2021 / 0147904.

[0498] Localization of TARGET and STANDARD peptide constructs on BINDERS produces a remarkable concentration effect. In some embodiments the BINDERS are attached to spherical magnetic particles, which can be gathered together into a compact mass by magnetic forces. In such a mass of spherical particles, the particles occupy about 74% of the total volume. Thus elution of constructs from binders on beads in such a compact mass releases the constructs into about 26% of the volume of the mass, and given the sub-microliter volume of this mass in many practical embodiments, the constructs can be present in volumes of interstitial liquid in the range of tens to hundreds of nanoliters.

[0499] In some embodiments, the immobilization of TARGET peptides, STANDARDS and their derivatives and constructs by non-covalent binding to BINDERS that are immobilized on supports such as magnetic beads, provide improved methods for delivery of these molecules to sequencing machinery. In some embodiments, magnetic beads carrying BINDERS and their peptide construct cargoes (including peptide-oligonucleotide constructs) are added directly to the cis chamber of a nanopore sequencing device, where the beads sink under the influence of gravity and come to lie on the membrane in which the nanopore is located. As described in US 2021 / 0147904, this presentation of sequenceable polymers adjacent to the membrane improves capture by the pore by orders of magnitude. In some embodiments, the salt solution of the cis compartment acts to slowly release the peptides from the BINDERS, for example with an off-rate equivalent to elution over a period of 15 to 180 minutes. Eluted peptide constructs are then captured by the membrane through a hydrophobic (e.g., cholesterol) tether as described in the commercial Oxford Nanopore device. Upon capture by the membrane, the peptide constructs diffuse in 2-dimensions on the membrane and are efficiently presented to the pore for capture and threading into the pore for sequencing.

[0500] In some embodiments, the BINDERS are attached to magnetic beads (or other particles) by a cleavable link such as a disulfide-containing linker. Exposure to a disulfide reducer (e.g., TCEP or mercaptoethanol) can thus release the BINDERS and their cargo from the beads into solution. By further providing the BINDER with a hydrophobic tether on a linker, the BINDER can be captured by the membrane and then diffuse in 2-dimensions, eventually bringing bound peptide constructs into close proximity to the nanopore.

[0501] In some embodiments, the BINDERS are released free into solution and the peptide constructs are not eluted from the BINDERS under the conditions of the cis chamber (e.g., 0.4M KC1), so that peptide constructs, still bound by BINDERS, are captured by and threaded into the nanopore. In some embodiments the force with which the electric field acts on the construct to pull it through the nanopore causes the construct to be pulled free of the BINDER while the speed of this motion is regulated by a DNA motor acting on the construct’s oligonucleotide component. Use of the pore forces to strip peptide constructs off of the cognate BINDER allows the BINDER to be used as a “chaperone” for the peptide construct throughout the journey from the peptide’s capture from a sample digest, though its modification to create a sequenceable construct, all the way to threading and delivery into a nanopore for sequencing.

[0502] In some embodiments, tethered versions of the BINDERS are added into the cis chamber and allowed to contact and be retained by the membrane, forming a dense surface of BINDER binding sites on the membrane within which the nanopore lies. This surface, comprising a dense plane of binding sites for TARGET and STANDARD peptide molecules, is able to capture these molecules from the contents of the cis compartment and thus provide an increased local concentration of constructs in the plane of the pore, and with the ability to diffuse in the membrane plane so as to deliver constructs for threading into the pore.

[0503] In some embodiments, peptide:oligonucleotide constructs can be propelled back and forth though a nanopore by reversal of the transmembrane electric potential (a process described as “flossing”) to repeatedly read and re-read a sequence to provide greater accuracy (65). In some embodiments of the present invention, this flossing approach is used to read selected peptides multiple times under computer control. This approach is particularly useful in confirming the sequences of peptides present at low abundance; i.e., when few copies of the peptide have been encountered and the potential for quantitative error is large. Thus, if a low frequency peptide is detected on its first pass through a nanopore, the nanopore control system can act on the observation of low frequency in real-time and implement a multi-read flossing protocol to verify the identity as a rare sequence. Achieving certain identification of a low frequency peptide sequence is more consequential than for high frequency peptides. 7.5.8 Inclusion of separative steps in addition to specific affinity capture and enrichment in a workflow.

[0504] In some embodiments additional separative steps are added to the workflow to improve performance. In some embodiments proteolytic peptides in a sample digest are captured on a solid support (e.g., C18-coated magnetic beads), thus allowing non-peptide sample components to be removed, allowing the immobilized peptides to be reacted with chemical reagents (e.g., for click chemistry derivatization as described below), and excess reagents to be removed from the peptides prior to their elution (e.g., using 50% acetonitrile) for use in subsequent steps. Likewise a conventional magnetic bead-based DNA “cleanup” (e.g., using commercially-available Ampure beads (Beckman Coulter)) can be carried out after assembly of oligo:peptide constructs generated according to the invention in order to remove any excess reagents, short oligos, un-linked peptides, and / or enzymes prior to delivery of peptide constructs to a single molecule sequencer (e.g., a nanopore). Modern laboratory robotics provides means to automate workflows involving multiple such affinity separative steps as well as precise liquid additions.

[0505] 7.6 USE OF THE INVENTION WITH SINGLE MOLECULE DETECTION BY NANOPORES.

[0506] In some embodiments a TARGET or STANDARD peptide is linked to a polymer at one or more sites (e.g., at one or both ends) by stable linkages (e.g., by covalent bonds or very stable non-covalent bonds). In some embodiments these linkages are made between specific sites at, or near, the peptide’s terminus (or termini) and one or more polymer molecules (e.g., nucleic acids including DNA or RNA, chemical variants of these including phosphorothioate backbones, “locked” nucleic acids (“LNA”), peptides including polyglutamic or polyaspartic acids, and the like). In many preferred embodiments, an object of the invention is to cause a peptide to pass through a nanopore in an extended conformation, allowing the sequence of amino acids to be “read” by measurement of current flowing through the pore or other equivalent means. In other embodiments, peptides of interest may be immobilized and subjected to a series of binding interactions or covalent modifications (e.g., stepwise degradation), and in such embodiments the linkage of peptides to a surface by a single unique site (e.g., a unique amino group such as an n-terminal amino group) may be preferred. 7.6.1 Peptide selection for pore signature and ligation properties.

[0507] Some embodiments make use of monitor peptides selected from among a target protein’s proteolytic fragments based on features including A) their ability to generate a distinct sequence or fingerprint in a single-molecule sequence-sensitive detector (e.g., a “squiggle” or ion current signature over time in a nanopore sequencing system) compared to other peptides (particularly other peptides that may be used with the selected peptide in a multiplex panel assay); B) their content of reactive groups (e.g., primary (amino terminal) and tertiary (lysine) amines, primary (carboxy terminal) and side chain (aspartic and glutamic acid) carboxyl groups, cysteine sulfhydryls, etc.) of potential use in labeling or ligation reactions; C) their uniqueness to a target protein (i.e., they are typically required to be “proteotypic” so as to act as a surrogate of the pre-specified target protein exclusively); and other properties desirable in a peptide analyte (e.g., solubility, chemical stability, etc.). Proteolysis using trypsin (i.e., cleaving polypeptides at lysine and arginine) is common in peptide analysis for several reasons including low cost of the enzyme and its ability to generate peptides having a positive charge at the c-terminus (a useful feature in mass spectrometric analysis). In the context of the present invention, selection from a tryptic digest of TARGET peptides with a c-terminal lysine ensures the presence of amino groups at both peptide termini, while selection of c-terminal arginine peptides ensures the presence of only one amino group (at the n-terminus). In the present invention alternative enzymes can be used as well. The enzyme LysC cuts only at lysines (not arginine) and therefore on average generates longer peptides than trypsin, and each of these (apart from a protein’s c-terminal peptide) will have a lysine (and its tertiary amino group) at the c-terminus (features that can be exploited for linking purposes in the invention). TARGET peptide selection criteria in the present invention are therefore significantly different from selection criteria for mass spectrometry.

[0508] 7.6.2 Double ligation strategy to attach Leader and Trailer.

[0509] Some embodiments apply a strategy to connect polymeric molecules (e.g., nucleic acids, polypeptides, and other polymers) to either or both ends of a ...

Claims

What is claim is:

1. A molecular construct and vehicle comprising:(a) a molecular construct comprising a peptide comprising a target peptide sequence derived from proteolytic cleavage of a target protein and a molecular tag defining the source of said peptide, and(b) a vehicle capable of presenting the construct for analysis by a sequence-sensitive single molecule detector.

2. The molecular construct and vehicle of claim 1, wherein the molecular tag is a target tag that identifies the peptide as a peptide created by proteolytic digestion of a biological sample.

3. The molecular construct and vehicle of claim 1, wherein the peptide comprises a synthetic peptide and the molecular tag is a standard tag that identifies the synthetic peptide as an internal standard.

4. The molecular construct and vehicle of claim 2, further comprising a sample barcode identifying the sample of origin.

5. The molecular construct and vehicle of claim 1 further comprising a binder barcode identifying a binder to which the construct has been bound.

6. The molecular construct and vehicle of claim 4, wherein the barcode or the tag is an oligonucleotide.

7. A standardized sample digest derived from a proteolytic digest of a biological sample, comprising:an amount of a molecular construct comprising a target tag and a target peptide, said construct being a target peptide construct and an amount of a molecular construct comprising a standard tag and a peptide whose sequence is the same or similar to the sequence of said target peptide, said construct being a standard peptide construct, wherein the target peptide is generated by proteolytic digestion of a target protein in said biological sample, wherein said target and standard tags can be distinguished by a single molecule detector and comprise chemical or structural groups covalently joined to peptides in their respective constructs, wherein said target tag is covalently attached to a plurality of the peptides present in said sample digest, wherein said target peptide construct comprises more than 90% of the target peptide molecules present in said sample digest and wherein said standard peptide construct is prepared separately and added to said digest in a known amount, or in a consistent relative amount across a multiplicity of samples. The standardized sample digest of claim 7, wherein the number of molecules of the standard peptide construct added to the sample digest differs by no more than a factor of 100 from the number of molecules of the target peptide construct in said sample digest. The standardized sample digest of claim 7, further comprising one or more additional standard peptide constructs having a different standard tag from each other and with each construct at a different relative abundance. The standardized sample digest of claim 7, wherein the target tag is covalently attached to a majority of the peptides generated by proteolytic digestion of said sample. The standardized sample digest of claim 7 wherein said tags are oligonucleotides.An enriched standardized sample digest, comprising a bound fraction of the standardized sample digest of claim 7 bound by a binder, wherein said bound fraction comprises a target peptide construct and a standard peptide construct in a ratio equal within 2%, 5%, 10% or 20% to the ratio in which they are present in said standardized sample digest. A stoichiometrically-flattened standardized sample, comprising a plurality of pairs of cognate standard and target peptide constructs enriched from a standardized proteolytic digest of a biological sample by binding to their respective cognate binders, wherein a pre-enrichment ratio calculated by dividing the number of molecules of a first target peptide construct that is the most numerous of said target peptide constructs in the standardized sample digest by the number of molecules of a second target peptide construct that is the least numerous of said target peptide constructs in the standardized sample digest is more than 10 times larger than a post-enrichment ratio calculated by dividing the number of molecules of said first target peptide construct by the number of molecules of said second target peptide construct in said enriched sample. A method for the measuring the amount of a selected target protein in a biological sample, comprising: proteolytically digesting said sample, modifying a plurality of peptides in the digested sample by adding a target tag to form a plurality of constructs comprising a selected target peptide derived from, and proteotypic of, said target protein, said plurality of constructs being target construct molecules, adding an amount that is known and / or consistent between a set of samples of a prepared standard peptide construct that is a cognate of said selected target peptide construct and comprises a standard tag, forming a standardized digest,enriching said cognate target and standard peptide constructs by contacting said standardized digest with a cognate binder, forming bound constructs, separating said bound constructs from unbound constructs to form enriched constructs, releasing said enriched constructs from said binder, linking said enriched constructs to a vehicle capable of presenting said enriched constructs to a sequence-sensitive single molecule detector, counting said enriched target construct molecules and said enriched standard construct molecules using a sequence-sensitive single molecule detector capable of distinguishing said target and standard tags and identifying said peptides, calculating the amount of said protein in said sample. The method of claim 14, wherein the calculating is performed by multiplying the amount of standard construct added by the ratio of the number of target construct molecules counted to the number of standard construct molecules counted by said detector. The method of claim 14, wherein said proteolytic digestion comprises at least two sequential steps resulting in peptide cleavage at different sites, and wherein peptides are covalently modified between two such steps (or wherein said first sequential step cleaves at lysine residues). The method of claim 14, wherein said proteolytic digestion comprises at least two sequential steps resulting in peptide cleavage at different sites, and wherein peptides retain an unmodified n-terminal amino group when presented to said detector. The method of claim 14, wherein a sample barcode is linked to said constructs encoding the identity, or relative position within a sample set, of said standardized samples; a plurality of said standardized samples is pooled; said sample barcodes associated with construct molecules are read using a sequence-sensitive single molecule detector; and the counts of target and standard construct molecules for each sample are separatedbased on said sample ID barcode identifying the sample from which they were enriched, and wherein said barcode may be an oligonucleotide. The method of claim 14, wherein a binder barcode is linked to said constructs identifying the binder by which they were enriched, and wherein said barcode may be an oligonucleotide. The method of claim 14, wherein said construct molecules are joined together into concatamers prior to presentation to said detector.

Images