Methods for the assembly of protein sequences in complex protein mixtures

Abstract

Antibodies are the main effectors of the adaptive immune system. Their unique characteristic of binding specific molecules known as antigens has generated many tools and strategies for diagnostics, research, and clinical applications. The ability to sequence several antibodies from a polyclonal mixture, or at least a few of its dominant forms (or a subgroup with specific binding characteristics), could potentially result in a faster procedure to generate recombinant antibodies. To date, a few efforts have been attempted to sequence polyclonal antibodies due to the complexity of the task. The present application relates to methods for assembling several complete antibody chains using a combination of methods (separation, biochemistry, and bioinformatics-based) including cross-linking, intact chain separation, middle- down proteomics and / or the use of combinatorial chain assembly databases with a proteomics search engine. The methods may further comprise recombinantly expressing the candidate antibodies identified to test their binding to a target antigen.

Description

METHODS FOR THE ASSEMBLY OF PROTEIN SEQUENCES IN COMPLEX PROTEIN MIXTURESCROSS REFERENCE TO RELATED APPLICATIONSThe present application claims the benefit of U.S. provisional patent application No. 63 / 517,222 filed on August 2, 2023, which is incorporated herein by reference.SEQUENCE LISTINGA sequence listing is submitted herewith as an XML file named G17294-00027_Seq Listing. xml, created on July 25, 2024, and having a size of 192,940 bytes. The content of the aforementioned file is hereby incorporated by reference in its entirety.TECHNICAL FIELDThe present invention generally relates to the field of antibodies, and more particularly to the identification and sequencing of antibodies in an antibody mixture.BACKGROUND ARTAntibodies are the main effectors of the adaptive immune system. Their unique characteristic of binding specific molecules known as antigens has generated many tools and strategies for diagnostics, research, and clinical applications. Antibodies are one of the fastest- developing biomolecules in clinical trials, with a worldwide market estimated at approximately 130.9 billion $ for 2020 and estimated to grow to 223.7 billion $ by the end of 20251.From a usage point of view, antibodies can be divided into three main categories: monoclonal antibodies (mAbs), which are produced by a single plasma cell and have the same heavy and light chain sequence and bind to a unique part of the antigen (i.e., epitope). On the other hand, polyclonal antibodies (pAbs) consist of a pool of antibodies from different B cells that recognize a mix of similar and different epitopes from the same antigen. Recombinant antibodies (rAbs) are monoclonal antibodies generated in vitro by cloning genes into an expression vector. These are then introduced into expression hosts to generate the associated antibody protein. Monoclonal antibodies are developed and used for several biological applications, including diagnostics and clinical applications, such as treatments for autoimmune diseases, cancer, and infectious diseases, to name a few. The next logical progression in immunotherapy is the combination of multiple monoclonal antibodies targeting distinct epitopes2. As the physiological immune response to an infection is more of a polyclonal response, it would be beneficial to develop oligoclonal / polyclonal immune therapy strategies.Choosing to produce and use mAbs or pAbs is influenced by different considerations, including application type, production cost and time, and technical expertise. Both types of antibodies have different advantages and disadvantages. Polyclonal antibodies can be generatedmore rapidly than mAbs with less demanding technical skills. All that is required is one or several animals being inoculated with a target antigen and adjuvant. In addition, pAbs are heterogeneous, which ensures they can recognize a given epitope under different conformations or with minor changes. In addition, pAbs are more flexible regarding buffer use and epitope conformational changes than mAbs. However, pAbs generated from different animals will exhibit significant variation in their affinity to a specific antigen. In addition, substantial variation in affinity within different bleeds from the same animal has been observed, thus making batch-to-batch variability a significant issue for polyclonal antibodies. One common strategy consists of mixing a large population of pAbs enriched from different animals to generate a product with an average performance that is most likely less prone to batch-to-batch variation. The resulting product is a mixture of excellent and mediocre antibodies, thus a relatively reproducible but average product.Unlike pAbs, mAbs are a homogeneous population of a single type and have low batch-to- batch variability, but their high specificity can sometimes limit their use. For example, a slight difference in the epitope structure or composition can significantly affect the antibody-antigen affinity. mAbs are generated from identical immortalized B-cells; fusing a B cell and a myeloma cell can sustain their production in vitro. However, hybridoma cells can be subjected to gene loss, gene mutation, and cell line genetic drift. These potential problems sometimes encountered in mAbs can be overcome using recombinant antibodies (rAbs). For recombinant antibody rAbs, the antibody sequence must first be determined to synthesize the immunoglobulin (Ig) L and H chain genes and generate expression constructs. The constructs are then transfected into a high-yield cell line such as CHO or HEK293.A similar strategy could be applied to pAbs. The ability to sequence several antibodies from a polyclonal mixture, or at least a few of its dominant forms (or a subgroup with specific binding characteristics), could potentially result in a faster procedure to generate recombinant antibodies of high quality. In addition, generating selected recombinant forms to produce a more straightforward complex mixture, such as a recombinant oligoclonal mixture, is an attractive solution, which combines advantages from both the mAbs and rAbs side (removal of batch-to- batch variability, circumvention of loss of hybridoma), and pAbs (a response closer to that of the natural immune system).To date, a few efforts have been attempted to sequence pAbs, most likely due to the complexity of the task.Cheung et al.3proposed a sequencing approach where antibodies were first enriched from an immunized animal and analyzed using a standard proteomics mass spectrometry (MS)-based approach searched against a reference database created by Next Generation Sequencing (NGS) of the B cell Ig repertoire of the immunized animal. Wine et al.4used a relatively similar approach. Gilchuk et al.5published an example of combining single-cell mRNA sequencing and proteomics to discover antibodies from human blood. Very little work has been done to sequence de novofrom a protein polyclonal mix. De novo polyclonal antibody sequencing can be advantageous, as having access to the B cell of the same animal used to generate the pAbs is often impossible. Moreover, NGS of the B cell repertoire applied to animals other than humans and mice is more explorative and requires some non-routine efforts in developing and testing proper primers.The three main challenges that need to be overcome to assemble few pure recombinant antibodies from either an artificial mixture or a natural polyclonal mixture are the following:Challenge 1 : Correctly assemble the different complementarity determining region CDRs Challenge 2: Correctly assemble independent heavy (H) and light (L) chains.Challenge 3: Correctly pair a given L chain with a given H chain.Assembling proteins from a peptide lysate of complex protein mixtures has been the subject of several studies and is often referred to as an inference problem6. This inference problem is encountered under specific experimental conditions such as the shotgun bottom-up proteomic approach. In this case, the complex protein mixture is initially subjected to digestion by single or multiple proteases; then, the peptides are separated and analyzed by LC-MS. The initial proteins can be identified by matching those peptide fragments to a database of known proteins. Under standard proteomics conditions, the sample mixture is quite heterogeneous, with only very few similar proteins which could share similar peptides. Although challenging, the problem is being solved in most standard proteomics facilities. The major challenge of assembling de novo individual relatively similar antibodies from a polyclonal mix is a significantly more complex one, which, to our knowledge, has not been performed routinely in any proteomics laboratory: (1) most of the present proteins share very similar sequence segments, and (2) there is an additional challenge that some sequence regions are more variable with no equivalent found in genomic databases and require using a de novo approach which does not rely on possible prior knowledge of the constituents present in the mixture (i.e., no match to a sequence database).Guthals et al.7attempted to sequence antibodies from a mixture derived from human serum. The authors managed to sequence several H and L chains and concentrated their efforts on the highest confident four LC and seven HC. Out of 28 antibodies expressed, only two showed affinities to the antigen. Notably, the affinities of these two antibodies are orders of magnitude lower than the polyclonal found in the original serum, indicating possible errors in their sequences. Their approach consists of generating a given chain (heavy or light) using a methodology based on assembling close, overlapping peptides and extending those reads. The overlapping peptides were generated by submitting the polyclonal mixture to different proteases and using an LC-MS proteomics-based approach to separate the peptide and generate fragments sequenced de novo. Those de novo sequenced peptides are assembled to generate consensus reads merged using overlapping “contigs” by combining the information from those short reads into a long stretch. Constructing long sequences based on a singular approach relies heavily on computing power and is prone to long-distance assembling errors. With such an approach, based on vicinityassembling, the confidence in correctly assembling two regions decreases rapidly along with the increasing distance between the two regions.More specifically, let us assume a protein contains n sequential regions, R_1 , R_2, ... , R_n, along its chain. In a complex protein mixture, there may be multiple choices for the sequence of each region. To correctly assemble the protein, the de novo sequencing process must ensure that all regions’ sequences come from the same protein. Based on the vicinity assembly approach, assume the confidence R_k and R_(k+1) belong to the same protein is “p”. Then the probability that all the n regions belong to the same protein drops exponentially to pA(n-1).For example, if p=0.9, the probability of correctly assembling 8 regions is only 0.9A7~0.478. In practice, the confidence of assembling two adjacent regions can be much lower. For example, when p=0.5, the probability of correctly assembling 8 regions is about 0.5A7~0.008, which is less than 1%. For this reason, exclusive usage of the vicinity approach to build long chains of antibodies is, therefore, severely limited. FIG. 1 illustrates graphically the limitation of assembling a long chain using only the neighbour approach.There is thus a need for novel methods to correctly assemble several complete antibody chains from complex mixtures.The present description refers to a number of documents, the content of which is herein incorporated by reference in their entirety.SUMMARYIn various aspects and embodiments, the present disclosure provides the following items 1 to 28:1 . A method for determining the amino acid sequence of one or more antibodies or antibody chains present in an antibody mixture, the method comprising:(A) generating a pool of antibody-derived peptide sequences by:(i) contacting a plurality of samples from the mixture with a reducing agent;(ii) optionally contacting the plurality of samples with an agent that either prevent disulfide bridge formation or modifies cysteine residues into lysine analogs;(iii) contacting the plurality of samples with one or more proteases and / or chemical proteolytic agents to obtain antibody-derived peptides, wherein each of the sample is contacted with a different protease, chemical proteolytic agent or combination thereof, thereby obtain a plurality of antibody-derived peptide digests;(iv) determining both the amino acid sequences and intensity of the short antibody- derived peptides present in the antibody-derived peptide digests by mass spectrometry using a de novo sequencing approach, wherein the short antibody- derived peptides have a length of less than 50, preferably less than 40, 30, or 20 amino acids;(v) assigning the antibody-derived peptide sequences to a specific complementary determining region (CDR1 , CDR2 or CDR3) or framework region (FR1 , FR2. FR3 or FR4);(vi) generating a library of candidate antibody chain sequences in silica by combining the antibody-derived peptides sequences from different antibody regions identified in (v);(B) performing at least one of (a) to (e) to identify the amino acid sequence of one or more antibodies or antibody chains present in the antibody mixture from the library of candidate antibody chain sequences:(a)(i) optionally contacting the sample with an immunoglobulin (Ig) domain isolation protease to obtain Fab, F(ab’)2 and Fc fragments; optionally removing the Fc fragments (e.g., using protein A / G beads);(ii) submitting the sample from the mixture to a separation step to separate the antibodies, antibody chains or antibody fragments Fab, (F(ab’)2 and Fc) present in the mixture in a plurality of fractions;(iii) optionally contacting a sample from the antibody mixture with a reducing agent and / or with an agent that modifies cysteine residues either to prevent disulfide bridge formation or to modify them into lysine analogs;(iv) contacting the fractions with or more proteases and / or chemical proteolytic agents to obtain digested fractions comprising antibody-derived peptides, wherein the peptides correspond to different regions of the antibodies or antibody chains;(v) submitting the digested fractions to mass spectrometry (MS) to obtain MS and / or MS / MS spectra;(vi) analyzing the MS and / or MS / MS spectra with a proteomic search engine using the pool of antibody-derived peptide sequences obtained in (A); and(vii) assembling the antibody-derived peptides issued from the same antibody or antibody chain based on their co-elution profile across the different fractions;(b)(i) optionally incubating a sample from the antibody mixture with an agent that modifies free cysteine residues;(ii) incubating the sample under denaturating, non-reducing conditions;(iii) contacting the sample with one or more proteases and / or chemical proteolytic agents to obtain digested antibody-derived peptides, wherein the digested antibody- derived peptides comprises dimers of peptides from distant antibody regions linked by a disulfide bridge;(iv) submitting the digested fractions to mass spectrometry (MS) to obtain MS and / or MS / MS spectra;(v) analyzing the MS and / or MS / MS spectra with a proteomic search engine suitable for identifying cross-linked peptides using the antibody-derived peptide sequences identified in (A);(c)(i) incubating a sample from the antibody mixture with a protein cross-linking agent;(ii) incubating the sample under denaturating, reducing conditions;(iii) optionally contacting the sample with an agent that modifies cysteine residues either to prevent disulfide bridge formation or to convert cysteine into lysine analogs;(iv) contacting the sample with one or more proteases and / or chemical proteolytic agents to obtain digested antibody-derived peptides, wherein the digested antibody- derived peptides comprise cross-linked dimers of peptides from distant antibody regions;(v) submitting the digested fractions to mass spectrometry (MS) to obtain MS and / or MS / MS spectra;(vi) analyzing the MS and / or MS / MS spectra with a proteomic search engine suitable for identifying cross-linked peptides using the antibody-derived peptide sequences identified in (A);(d)(i) generating a library of candidate antibody chain sequences in silica by combining the antibody-derived peptides sequences from different antibody regions identified in (A)(v);(ii) determining, by mass spectrometry, both the amino acid sequences and intensity of the long antibody-derived peptides present in the plurality of antibody-derived peptide digests from (A)(iii) , wherein the long antibody-derived peptides have a length of more than 20, preferably more than 30, 40 or 50 amino acids;(iii) comparing the long antibody-derived peptide sequences with the library of candidate antibody chain sequences to identify antibody chain sequences present in the antibody mixture;(e)(i) assessing the overlap between the antibody-derived peptides present in the sample, wherein a high level of overlap is indicative that the antibody-derived peptides belong to the same antibody or antibody chain.2. The method of item 1 , wherein the separation step comprises a chromatography or gel separation.3. The method of item 2, wherein the chromatography is Hydrophobic Interaction Chromatography (HIC), and wherein the gel separation is native gel, isoelectric focusing (IEF) gel, or 2D gel separation.4. The method of any one of items 1 to 3, wherein the separation step is performed under nonreducing conditions.5. The method of any one of items 1 to 4, wherein the protease is pepsin, trypsin, chymotrypsin, AspN, LysC, GluC or any combinations thereof.6. The method of any one of items 1 to 5, comprising incubating the sample from the antibody mixture with a protein cross-linking agent.7. The method of item 6, wherein the protein cross-linking agent comprises bis(sulfosuccinimidyl)suberate (BS3).8. The method of any one of items 1 to 7, wherein the protein sequence database search engine is Mascot, Sequest, Novor-Cloud or Maxquant.9. The method of any one of items 1 to 8, which comprises contacting the sample with an agent that modifies cysteine residues into lysine analogs.10. The method of item 9, wherein the agent that modifies cysteine residues to prevent disulfide bridge formation comprises iodoacetamide, and / or the agent that modifies cysteine residues into lysine analogs comprises an electrophilic ethylamine, for example 2-Bromoethylamine hydrobromide (BEA).11. The method of any one of items 1 to 10, which comprises contacting the sample with an agent that prevents cysteine residues from forming disulfide bonds.12. The method of item 11 , wherein the agent that prevents cysteine residues from forming disulfide bonds comprises N-ethylmaleimide.13. The method of any one of items 1 to 12, wherein the Ig domain isolation protease comprises IdeS and / or IdeZ.14. The method of any one of items 1 to 13, wherein the short antibody-derived peptides have a length of 5 to 20, 30, 40 or 50 amino acids.15. The method of any one of items 1 to 14, wherein the long antibody-derived peptides have a length of 20, 30, 40 or 50 to 100 amino acids.16. The method of item 15, wherein the long antibody-derived peptides have a length of 40 to 80 amino acids.17. The method of any one of items 1 to 16, wherein the level of overlap between two antibody- derived peptides is determined by calculating an overlap score, and wherein an overlap between amino acids located at the amino and carboxy-terminal ends of the antibody-derived peptides is given a lower score relative to an overlap between amino acids located at internal positions in the antibody-derived peptides.18. The method of any one of claims 1 to 17, wherein the method comprises performing at least(a).19. The method of any one of claims 1 to 18, wherein the method comprises performing at least(b).20. The method of any one of claims 1 to 19, wherein the method comprises performing at least(c).21 . The method of any one of claims 1 to 20, wherein the method comprises performing at least(d).22. The method of any one of claims 1 to 21 , wherein the method comprises performing at least(e).23. The method of any one of items 1 to 22, wherein the antibody mixture is a polyclonal antibody mixture or a mixture of monoclonal antibodies.24. The method of any one of items 1 to 23, wherein the MS is liquid chromatography-MS (LC- MS) and the MS / MS is tandem MS.25. The method of any one of items 1 to 24, further comprising expressing one or more recombinant antibodies or antibody fragments corresponding to the one or more antibodies identified by the method defined in any one of items 1 to 14.26. The method of item 25, further comprising assessing the binding of the recombinant antibody or antibody fragment to a target antigen.27. The method of item 26, wherein assessing the binding of the antibody or antibody fragment comprises performing an immunoassay.28. The method of item 27, wherein the immunoassay is enzyme-linked immunosorbent assay (ELISA).Other objects, advantages and features of the present disclosure will become more apparent upon reading of the following non-restrictive description of specific embodiments thereof, given by way of example only with reference to the accompanying drawings.BRIEF DESCRIPTION OF DRAWINGSIn the appended drawings:FIGs. 1A-B are graphs depicts the confidence in pairing two regions in the function of the distance separating them. This figure is an approximation to illustrate that the confidence that two regions are part of the same chain will decrease as they are far apart. FIG. 1A: assuming a homogenous confidence of 90% confidence that two adjacent specific peptides are in the same chain, this confidence drops to 59% for R1 and R6 and decreases to around 25% for R1 and R14. FIG. 1 B: the influence of the probability score from 90%, 80%, and 50% for two neighbors to be parts of the same chains and the distance between a given sequence and any other sequence within a chain.FIG. 2A shows a description of the used nomenclature to illustrate the naming of the region (CDRs and framework region for both light and heavy chains).FIG. 2B depicts the process of assembling regions following peptide digestions. Using different proteases (with different amino acid specificity) on the same sample allows for generating short sequences read that can overlap (an important step for a proper assembly).FIG. 2C shows an example of distant regions of a given heavy antibody chain in a mixture of different antibodies. Pairing the proper CDRs can be done pair-wise (2.1 , 2.2, and 2.3) and verified for the global association of the specific CDR1-CDR2-CDR3 (2.4).FIG. 3A depicts a distant region being covalently connected through either a natural covalent bond (disulfide bridge) or artificially through a crosslinker (through lysine, for example). CDR1 and CDR3 are “distant” regions from a linear point of view. However, they are covalently connected through a disulfide bridge or can be covalently connected through a crosslinker, thus increasing the confidence in associating those two distant regions.FIG. 3B is a schematic illustrating that using only a “vicinity approach” (VA) to build an entire chain just leads to lower confidence the farther we are in the chain. Confidence can be increased here by using a disulfide bridge or crosslink to identify the connection between CDR1 and CDR3, affecting global confidence in the entire chain sequencing. Black = high confidence while white illustrates low confidence.FIG. 4 depicts a specific example of a given germline rabbit antibody (IGHV1S1 see IMGT) with cysteine (C) and lysine (K) being highlighted. The CDR1 and CDR3 have been underlined. If such an antibody is digested with lysC (no reduction), two peptides linked together by their disulfide bridge that contain a specific CDR1 and CDR3 should be obtained. This given peptide can be analyzed “as is” using a middle-down proteomics method or separated in gel or followed by another protease (e.g., pepsin, gluC, or AspN), in this example, lysC followed by pepsin digestion is shown. Here the application is to associate a specific CDR1 with a specific CDR3 in a mixture of other antibodies and increase the confidence in assembling distant regions within a given antibody.FIG. 5A depicts a generic procedure to assemble a given antibody in the presence of other antibodies. In Step 1 : a polyclonal mix is digested with different proteases. In step 2, those peptides are converted into sequence information. Step 3, Those regions are combined using a combinatorial approach in Fasta format. Those fragments can either be entire antibodies or Fab regions to simply shorter reads (few FRs and CDRs assembled). This combinatorial database can be used to search the initial dataset generated (step 1). Or a subset of it can be used to search specific experimental datasets (step 4) and search using a smaller dataset than step 3. This process steps 3 to 5, can be repeated a few times under different experimental conditions. In the end, those regions are assembled into chains, and can be expressed recombinantly, and their affinity tested in a functional assay.FIG. 5B: generation of a list of possible regions using a de novo approach can be performed using standard approaches such as the software Novor and Peaks (Bioinformatics Solutions). Some of the germlines and more conserved regions can be determined using standard search engines with the use of publicly available sequence databases (imgt.org, Uniprot, GenBank, RefSeq).FIG. 5C: Those regions can be assembled using a combinatorial approach and used as a database for a proteomics search engine. The main scoring strategy (i.e., ranking possible hits and discarding noise) is performed by using an overlap score (instead of total peptide identified), briefly, a confidence score is weighted toward the quality of the peptide overlap sequences instead of total peptide identified, only the sequences showing a high level of overlap are kept, and the rest of the hits are discarded (see FIG. 8). This process can be performed in a modular manner.FIG. 5D: series of experiments allow narrowing down the list of possible candidates to only a very number and reducing the number of false positive.FIG. 6 depicts HIC-HPLC separation and fractionation of a mixture of 2 mouse antibodies from Absolute Antibody (referred to as “P13” and “P14”). The four-rectangle area highlights four fractions that were reduced, digested, and analyzed by LC-MS.FIG. 7 depicts the sequence naming strategy used in this work and how the link between the sequence name and the sequence itself exists. In this example, the region data for the alpaca antibody of Example 2 is used. Table 3 illustrates different possible FR1 , CDR1 , FR2, etc, they can be arranged in a matrix structure or vector. Based on Table 3, the vector FR1 [1] is the sequence EVHLVESGGGLVQPGGSLRLSCWS (i.e., the first sequence in that table) and so on. The name “seq-1 -2-3-1 -2-7-1” refer then to the assembly FR1 [1]-CDR1 [2]-FR2[3]-CDR2[1]- FR3[2]-CDR3[7]-FR4[1], The letters “a,” “b,” and “c” are used to design any possible combination (i.e. “a” for FR1 mean any of the five possible FR1 sequences can be used from table 3) or a specific number for a specific sequence (i.e., seq-1 -b-c-d-e-f-g mean FR[1] and the rest of the fab region is any possible combination), finally, either a limited range of combination can also be used seq-1 -(2-3)-c-d-e-f-g (i.e., FR1 [1] then either CDR1 [2] or CDR1 [3], etc.). Such a naming allows the generation of a combinatorial sequence rapidly or to obtain a list of possible sequences. In addition, analyzing a selected possible group of sequences obtained after a database search permits to see a possible emerging pattern (all of the possible sequences have the same FR1 , CDR1 , etc.).FIG. 8 shows that typical search engines like Novor-cloud (or any other possible search engine such as Mascot, Sequest, GPM, etc.) rank and score proteins based on the number of possible experimental peptides supporting a given sequence. In this present work, an additional “quality score” was developed, which can be applied especially in the case of the use of multiple different protease-generating overlapping peptides. Each peptide matching a given proteinsequence is transformed into a vector of the same length in amino acids and carries a discrete value from 0 (N and C terminal extremity) to a maximum value of 3 by increment (each step farther from the N and C extremity). This numeric vector is positioned at the same position as the identified peptide in the sequences. A summation across all peptides is then performed from the protein N-term to the C-terminal end of the protein. In the presented case, although case A shows more peptides to explain the FASTA sequence “case A” (i.e., it will most likely be ranked high by any search engine). However, the minimal overlap is “0” while case B, with significantly less peptide shows a minimum score of 1 . Thus, using simply any search engine to score protein rank (and based on total peptides identified), case A will be ranked as the most probable case (i.e. as more total peptides identified), but using the strategy develop in this work based on the minimum overlap score, case B is the one that is the most probable (total number of peptides is less than A but the quality of the overlap peptide is higher).FIG. 9A depicts an overlap score plot for the best 4 top candidates for the unique CDR3[7], All of the four sequences are similar from FR1 to CDR3 (i.e. seq- 1-2-3- 1-2-7) as shown in the upper part of the figure and differ only by their FR4 / J region and hinge region. A similar coverage of the entire sequence (FR1 to FR3) can be seen, and it may be noticed that the C-term side of the CDR3 is better covered with FR4[1] and hinge[2] i.e., G03).FIG. 9B shows an SDS page gel performed under non-reduced conditions of an immunoprecipitation experiment with a rabbit IgG as the antigen and using the natural alpaca polyclonal. Several bands were identified, one above 150kDa, which represents the canonical alpaca lgG1a,b, and a band at 15kDa that is a fragment of the non-canonical form lgG2b,c that is naturally found in the sample.FIG. 10A shows a heatmap of the top 10 candidate sequences. White is associated with low score overlap, and the darker area is associated with high score overlap. A better overlap score is observed with CDR3

[0010] , Four sequences containing the CDR3

[0010] are highlighted.FIG. 10B is a plot of the overlap score for the top 4 selected sequences with the selected CDR3

[0010] ,FIG. 11A depicts the structure of the recombinant VHH antibody. The VHH domain was made as a dimer with a human Fc lgG1 and the human hinge region.FIG. 11B depicts an ELISA plot of the two recombinants presented in the present studies with two negative controls. The natural alpaca anti-IgG rabbit cannot be plotted in the same ELISA as the two recombinants that were made with human Fc, thus requiring different secondary antibodies, although a midpoint in the curve is observed around 0.2-0.4 nM.FIGs. 12A-B show the distribution of the minimum score for the combinatorial heavy (FIG. 12A) for the CDR3H[3] and light (FIG. 12B) for the CDR3 10] chain of a rabbit antibody.FIG. 13 shows the overlap score for the top 3 best heavy chain sequences identified after nonreduction digestion and combinatorial database search (rabbit polyclonal). They mostly differ by their CDR3 and FR4 chains.FIG. 14 shows the overlap score for the top 3 best light chain sequences identified after nonreduction digestion and combinatorial database search. They mostly differ by their FR2 and CDR2 regions.FIG. 15 shows an ELISA plot of the recombinant antibody PD108_R1 compared to the total polyclonal mix both with mid-point in the low nM range.FIG. 16 shows the native gel separation of the natural polyclonal human antibody anti-RBD isolated from a vaccinated individual. The polyclonal antibody was digested with IdeS to generate fab2fragments and an Fc fragment.FIGs. 17A-C are panels representing seven different contigs for heavy CDR1 (FIG. 17A), CDR2 (FIG. 17B) and CDR3 (FIG. 17C). In each panel (i.e., representing a given contig), the normalized intensity of the different peptides part of a given contig are then plotted.FIG. 18 shows the three contigs obtained and selected, aligned to the germline gene. After filling the gaps with additional overlapping de novo peptides and correcting the He and Leu, a final sequence of the full heavy chain variable region was obtained.DETAILED DISCLOSUREThe use of the terms “a” and “an” and “the” and similar referents in the context of describing the technology (especially in the context of the following claims) are to be construed to cover both the singular and the plural, unless otherwise indicated herein or clearly contradicted by context.The terms “comprising”, “having”, “including”, and “containing” are to be construed as open- ended terms (i.e., meaning “including, but not limited to”) unless otherwise noted.All methods described herein can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context.The use of any and all examples, or exemplary language (“e.g.”, “such as”) provided herein, is intended merely to better illustrate embodiments of the claimed technology and does not pose a limitation on the scope unless otherwise claimed.No language in the specification should be construed as indicating any non-claimed element as essential to the practice of embodiments of the claimed technology.Herein, the term “about” has its ordinary meaning. The term “about” is used to indicate that a value includes an inherent variation of error for the device or the method being employed to determine the value, or encompass values close to the recited values, for example within 10% of the recited values (or range of values).Recitation of ranges of values herein are merely intended to serve as a shorthand method of referring individually to each separate value falling within the range, unless otherwise indicatedherein, and each separate value is incorporated into the specification as if it were individually recited herein. All subsets of values within the ranges are also incorporated into the specification as if they were individually recited herein.Where features or aspects of the disclosure are described in terms of Markush groups or list of alternatives, those skilled in the art will recognize that the disclosure is also thereby described in terms of any individual member, or subgroup of members, of the Markush group or list of alternatives.Unless specifically defined otherwise, all technical and scientific terms used herein shall be taken to have the same meaning as commonly understood by one of ordinary skill in the art (e.g., in stem cell biology, cell culture, molecular genetics, immunology, immunohistochemistry, protein chemistry, and biochemistry).Unless otherwise indicated, the recombinant protein, cell culture, and immunological techniques utilized in the present disclosure are standard procedures, well known to those skilled in the art. Such techniques are described and explained throughout the literature in sources such as, J. Perbal, A Practical Guide to Molecular Cloning, John Wiley and Sons (1984), J. Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbour Laboratory Press (1989), T. A. Brown (editor), Essential Molecular Biology: A Practical Approach, Volumes 1 and 2, IRL Press (1991), D. M. Glover and B. D. Hames (editors), DNA Cloning: A Practical Approach, Volumes 1- 4, IRL Press (1995 and 1996), and F. M. Ausubel et al. (editors), Current Protocols in Molecular Biology, Greene Pub. Associates and Wiley-lnterscience (1988, including all updates until present), Ed Harlow and David Lane (editors) Antibodies: A Laboratory Manual, Cold Spring Harbour Laboratory, (1988), and J. E. Coligan et al. (editors) Current Protocols in Immunology, John Wiley & Sons (including all updates until present).All presented regions from antibodies in this document are based on the Chothia numbering scheme.The studies reported herein demonstrate that to correctly assemble single or multiple chains, the chain assembly based on the vicinity approach can be significantly improved by incorporating several pieces of complementary experimental evidence. These pieces of evidence increase confidence in the proper pairing of distant regions.In a first aspect, the present disclosure provides a method for determining the amino acid sequence of one or more antibodies or antibody chains present in an antibody mixture, the method comprising:(A) generating a pool of antibody-derived peptide sequences by:(i) contacting a plurality of samples from the mixture with a reducing agent;(ii) optionally contacting the plurality of samples with an agent that either prevent disulfide bridge formation or modifies cysteine residues into lysine analogs;(iii) contacting the plurality of samples with one or more proteases and / or chemical proteolytic agents to obtain antibody-derived peptides, wherein each of the sample is contacted with a different protease, chemical proteolytic agent or combination thereof, thereby obtain a plurality of antibody-derived peptide digests;(iv) determining both the amino acid sequences and intensity of the short antibody- derived peptides present in the antibody-derived peptide digests by mass spectrometry using a de novo sequencing approach, wherein the short antibody- derived peptides have a length of less than 50, 40, 30, or 20 amino acids;(v) assigning the antibody-derived peptide sequences to a specific complementary determining region (CDR1 , CDR2 or CDR3) or framework region (FR1 , FR2. FR3 or FR4);(vi) generating a library of candidate antibody chain sequences in silica by combining the antibody-derived peptides sequences from different antibody regions identified in (v);(B) performing at least one of (a) to (e) to identify the amino acid sequence of one or more antibodies or antibody chains present in the antibody mixture from the library of candidate antibody chain sequences:(a)(i) optionally contacting the sample with an immunoglobulin (Ig) domain isolation protease to obtain Fab, F(ab’)2 and Fc fragments; optionally removing the Fc fragments (e.g., using protein A / G beads);(ii) submitting the sample from the mixture to a separation step to separate the antibodies, antibody chains or antibody fragments Fab, (F(ab’)2 and Fc) present in the mixture in a plurality of fractions;(iii) optionally contacting a sample from the antibody mixture with a reducing agent and / or with an agent that modifies cysteine residues either to prevent disulfide bridge formation or to modify them into lysine analogs;(iv) contacting the fractions with or more proteases and / or chemical proteolytic agents to obtain digested fractions comprising antibody-derived peptides, wherein the peptides correspond to different regions of the antibodies or antibody chains;(v) submitting the digested fractions to mass spectrometry (MS) to obtain MS and / or MS / MS spectra;(vi) analyzing the MS and / or MS / MS spectra with a proteomic search engine using the pool of antibody-derived peptide sequences obtained in (A); and(vii) assembling the antibody-derived peptides issued from the same antibody or antibody chain based on their co-elution profile across the different fractions;(b)(i) optionally incubating a sample from the antibody mixture with an agent that modifies free cysteine residues;(ii) incubating the sample under denaturating, non-reducing conditions;(iii) contacting the sample with one or more proteases and / or chemical proteolytic agents to obtain digested antibody-derived peptides, wherein the digested antibody- derived peptides comprises dimers of peptides from distant antibody regions linked by a disulfide bridge;(iv) submitting the digested fractions to mass spectrometry (MS) to obtain MS and / or MS / MS spectra;(v) analyzing the MS and / or MS / MS spectra with a proteomic search engine suitable for identifying cross-linked peptides using the antibody-derived peptide sequences identified in (A);(c)(i) incubating a sample from the antibody mixture with a protein cross-linking agent;(ii) incubating the sample under denaturating, reducing conditions;(iii) optionally contacting the sample with an agent that modifies cysteine residues either to prevent disulfide bridge formation or to convert cysteine into lysine analogs;(iv) contacting the sample with one or more proteases and / or chemical proteolytic agents to obtain digested antibody-derived peptides, wherein the digested antibody- derived peptides comprise cross-linked dimers of peptides from distant antibody regions;(v) submitting the digested fractions to mass spectrometry (MS) to obtain MS and / or MS / MS spectra;(vi) analyzing the MS and / or MS / MS spectra with a proteomic search engine suitable for identifying cross-linked peptides using the antibody-derived peptide sequences identified in (A);(d)(i) generating a library of candidate antibody chain sequences in silica by combining the antibody-derived peptides sequences from different antibody regions identified in (A)(v);(ii) determining, by mass spectrometry, both the amino acid sequences and intensity of the long antibody-derived peptides present in the plurality of antibody-derived peptide digests from (A)(iii) , wherein the long antibody-derived peptides have a length of more than 20, preferably more than 30, 40 or 50 amino acids;(iii) comparing the long antibody-derived peptide sequences with the library of candidate antibody chain sequences to identify antibody chain sequences present in the antibody mixture;(e)(i) assessing the overlap between the antibody-derived peptides present in the sample, wherein a high level of overlap is indicative that the antibody-derived peptides belong to the same antibody or antibody chain.In an embodiment, at least (a) is performed. In an embodiment, at least (b) is performed. In an embodiment, at least (c) is performed. In an embodiment, at least (d) is performed. In an embodiment, at least (e) is performed.In an embodiment, at least two of (a)-(e) are performed. In an embodiment, two of (a)-(e) are performed. In an embodiment, (a) and (b) are performed. In another embodiment, (a) and (c) are performed. In another embodiment, (a) and (d) are performed. In another embodiment, (a) and (e) are performed. In another embodiment, (b) and (c) are performed. In another embodiment,(b) and (d) are performed. In another embodiment, (b) and (e) are performed. In another embodiment, (c) and (d) are performed. In another embodiment, (c) and (e) are performed. In another embodiment, (d) and (e) are performed.In an embodiment, at least three of (a)-(e) are performed. In an embodiment, three of (a)- (e) are performed. In an embodiment, (a), (b) and (c) are performed. In an embodiment, (a), (b) and (d) are performed. In an embodiment, (a), (b) and (e) are performed. In an embodiment, (a),(c) and (d) are performed. In an embodiment, (a), (c) and (e) are performed. In an embodiment, (a), (d) and (e) are performed. In an embodiment, (b), (c) and (d) are performed. In an embodiment, (b), (c) and (e) are performed. In an embodiment, (c), (d) and (e) are performed.Close or vicinity regions refers to at least two adjacent peptides from an antibody, but can be extended to adjacent regions such as FR1 with CDR1 , CDR1 with FR2, FR2 with CDR2, and other configurations shown in FIG. 2A. Distant regions correspond to two regions separated by at least one additional region. The following eight examples are considered distant regions:CDR1 with CDR2,CDR2 with CDR3,CDR1 with CDR3,FR1 with FR2,FR2 with FR3,FR1 with FR3,FR1 with CDR2 or CDR3,FR2 with CDR3, etc.The following terminology is used herein for differentiating heavy and light chains: FR1 H and FR1 L. Those two examples correspond to framework 1 of the Heavy and Light chains, respectively. Assuming a mix of “i” different antibodies, the terminology can be extended as follows: FR1 Hi and FR1 Li, as framework 1 , Heavy and Light, respectively, of a given antibody “i.” A specific CDR1 H1 is part of the same chain as CDR2H1 and CDR3H1.It is important to remember that, in the case of antibodies, for single-chain assembly (either the heavy or light chain), the functionality of a given antibody relies heavily on correctly assembling three specific distant regions: the CDRs (CDR1 , CDR2, and CDR3), in addition to the correct heavy and light chain pairing. Although the framework regions (FRs) are subject to mutations, a loose combination can be defined based on germline usage. However, their exact composition is important for antibodies’ activity as well.The challenge that the present disclosure wants to address is to assemble several complete antibody chains using a combination of methods (separation, biochemistry, and bioinformaticsbased) by assembling correctly within a chain, specific CDR1 , CDR2, and CDR3, and to extend that to include FR1 , FR2, FR3, and FR4.FIGs. 2A-C demonstrates the nomenclature and assembly of individual regions and close regions being assembled, while FIG. 2C highlights the pairing of distant regions within a given chain.To improve the accuracy of those long stretches of sequence, It is proposed to utilize the combination of different approaches in addition to the vicinity assembling approach, presented above by Guthals et al.7. The proposed approaches are based on either unique structural characteristics inherent to each antibody or other experimental approaches, mainly based on protein separation methods and data analysis.These specific approaches fit into four different main categories:First category: The use of covalent bonds other than peptide bonds:This category includes the use of 1) natural disulfide bridges and 2) artificial crosslinks (e.g., using a protein cross-linking agent) as a strategy to assemble long-distance regions within a given sequence. With this approach, distant regions covalently linked due to their proximity in the antibody’s higher-order structure are detected. For example, in both heavy and light IgG chains, the first and second cysteine from the N-terminal end is often linked, allowing a loop to be formed. This loop includes the CDR1 , CDR2, and CDR3 (See FIG. 3). As shown in FIG. 3A, under nonreducing conditions, a specific peptide representing the FR1-CDR1 can be associated with its FR3-CDR3, similarly, through artificial crosslinking of a primary amine (i.e., lysine) followed by digestion with a specific protease or a protease combination, could also highlight specific longdistance correct region association. In FIG. 3B, the global confidence in assembling a chain using only a vicinity approach (VA) is compared to a combination of approaches such as the vicinity approach + disulfide bridge information. The confidence of pairing distant regions can be increased by using those additional targeted experiments. Using such an approach allows associating a unique FR1-CDR1 to a unique FR3-CDR3 in a mixture of other very similar FR1 , CDR1 , FR3, and CDR3.FIG. 4 illustrates the outcome of using specific proteases on the heavy rabbit chain from the germline IGHV1S1 under non-reducing conditions. Using the protease Lys-C under non-reducingconditions should allow isolating a long peptide composed of multiple regions, including a CDR1 paired with a CDR3 linked by a natural disulfide bridge. This peptide can then be analyzed directly by LC-MS or digested a second time with a different protease. It is then possible to isolate disulfide-linked peptides, which will include signature peptide sequences from both a specific CDR1 and from another specific CDR3, thus allowing a given unique CDR1 to be paired with a given unique CDR3 in a mixture of antibodies (in a mixture of several other CDR1 , CDR2 and CDR3 regions that are present). Artificial cross-linking also allows specific unique distant region pairing within an antibody. The selected example presented in the present disclosure is done between cross-linked lysine residues, but any suitable protein cross-linking agent may be used. The protein cross-linking agent may be a bifunctional reagent or heterofunctional reagent. The reactive function of the protein cross-linking agent can be an NHS ester compound (that reacts with primary amines present in the side chain of lysines), a maleimide compound (that reacts with sulfhydryl containing molecule present in the side chain of cysteines), a hydrazide compound (that reacts with aldehyde containing molecule), or a carbodiimide based compound such as EDC (that reacts with carboxylate containing molecule). The two functional groups may be separated by a spacer, such as an alkyl or polyethylene glycol (PEG) chain. Agents to induce protein cross-linking are well known in the art and include, for example, glutaraldehyde, DSG, disuccinimidyl suberate (DSS), Bis(sulfosuccinimidyl) suberate (BS3), Bis(succinimidyl) penta(ethylene glycol (BS(PEG)5), TSAT, DSP, DTSSP, DST, BSOCOES, EGS, Sulfo-EGS, DMA, DMP, DMS, DTBP, DFDNB, SIA, SMAP, SIAB, Sulfo-SIAB, AMAS, BMPS, GMBS, Sulfo-GMBS, MBS, Sulfo-MBS, SMCC, Sulfo-SMCC, SMBP, Sulfo-SMBP, SMPH, LC-SMCC, Sulfo-KMUS, SPDP, LC-SPDP, Sulfo-LC-SPDP, or SMPT. Analysis of cross-linked peptides may be performed using suitable tools such as pLink8, ECL9, xQuest10 11, ProteinProspector12 13, Kojak14, OpenPepXL15or MS- Annika16. For this approach, a search engine with suitable for proteome-scale identification of cross-linked peptides is used. Examples of such search engines include xQuest, StavroX, pLink / pLink 2, xProphet, Protein Prospector, Kojak, Xi, Xilmass, MetaMorpheusXL, and Xolik17. In an embodiment, the search engine is pLink / pLink 2.Second category: Intact Chain separation using any chemistry / biochemistry separation strategy (gel, Liquid chromatography, HPLC) and information clustering based on LC-MS peak intensity:In that category, chains are separated intact based on their physicochemical properties. The complex intact polyclonal mixture is then separated into multiple fractions, either intact antibody or after reduction, using HPLC, standard Liquid chromatography, gel (native or 2D gel electrophoresis), or any analytical separation strategy. Antibody domains can be as well separated (Fab / Fab2instead of intact antibody) by using specific proteases such as IdeS, IdeZ, papain, or even pepsin. These different fractions are then digested using single or multiple proteases and then analyzed with LC-MS. Distant regions from the same sequence can then beassembled based on their intensity profile across the different fractions. This can be done by correlating peak intensities, as distant regions of the same antibody will have similar elution profiles across different fractions.Third category: Middle-down proteomics: Longer fragments contain fragments from several regions within a single peptide, thus confirming a specific pairing of adjacent CDR regions (combination of MS / MS fragmentation with specific peptide mass). A combinatorial database of different CDRs and FRs is generated based on multiple protease digests and vicinity assembly. Experimental middle-down datasets are matched against the generated database to identify a possible combination of CDRs and FRs.Fourth category: Using a proteomics search engine with a Combinatorial Chain Assembly database and scoring the chain assembly quality. This approach relies on generating the different possible regions, then combining them exhaustively or limiting possible chain assembly using information deduced from experimental observations. The generated database is then used with a proteomics search engine. The selected possible candidate(s) is primarily based on the quality of the overall assembly (i.e., quality of the different peptides overlap obtained from using different proteases) and less on the total number of peptides identified for a given possible candidate sequence. Using a quality overlap instead of total peptide has not been often explored in the literature for proteomics analysis.None of the proposed approaches are universal due to the variability of the antibody sequence properties. Combining these several methods allows the confirmation of chain assemblies with higher confidence.In the studies reported in the Examples, two different strategies were used to validate the four approaches described above.Strategy 1: Use of known standard antibody mixtures. The antibody standards have been previously sequenced; therefore, correctly assembling distant regions can be easily checked against the standard known sequence.Strategy 2: Method validation by protein expression. Validation of the procedure(s) to identify a few unique sequences within a pool of natural polyclonal pAbs is possible through the recombinant expression of the sequenced forms. The recombinant antibodies are then tested for their affinity against the natural polyclonal antibody. The procedure can be validated by obtaining recombinant antibodies with high affinity against the same target from a de novo project.FIG. 5A presents a generic procedure to illustrate the approach for sequencing and validating a polyclonal antibody.Step 1 digestion by several proteases, either in series or in parallel, of a reduced and alkylated polyclonal antibody is completed. Peptide fragments are analyzed by LC-MS.Step 2. The MS and MS / MS data are used to generate the different regions i.e., CDRs, and FRs regions using a de novo sequencing approach using software such as Novor or Peaks (Bioinformatics Solutions). Some of the germlines and more conserved regions can be determined using standard search engines with the use of publicly available antibody sequences (from IMGT, Uniprot, GenBank, RefSeq, for example).Step 3. Those regions are assembled into sequences using a combinatorial approach and allow the construction of the FASTA database of combinatorial sequences.In Step 4, targeted experiments are performed to identify correct long-distance pairings, including, but not limited to, protease digestion under non-reducing conditions, protein separation, crosslink usage, and analysis of longer peptide fragments (middle-down proteomics).Step 5. Data analysis is performed using a database search engine. pLink or any software having the ability to perform analysis on crosslinked or non-reduced peptides18. For experiments not involving any non-reduced or crosslink peptide, Novor cloud, or any search engine based on sequence matching against a database (i.e., Mascot, Sequest, Maxquant, and similar search engines), may be used. The heavy and light chain pairing may be performed using different experimental conditions (see, e.g., Le Bihan et a / .19).In Step 6, based on those experimental and data analyses, a series of recombinant proteins are made and tested for their binding affinity. A more detailed process is presented in FIGs. 5B- D. Variations around this procedure were used, such as manual assembly of a few regions / contigs then, intact protein separation using native gel, distant region assembly based on elution profile, then gap filling using germline match.In FIG. 5B, the two main elements that are important for generating a list of possible candidate sequences are highlighted. These two elements are: 1) raw data (MS / MS) and 2) de novo search of those raw data to generate a list of a plurality of the different regions (frameworks and complementary determining regions). Those short reads can be obtained using any de novo software such as Novor cloud, Peak, etc., or manual de novo sequencing. Some of the germlines and more conserved regions can be determined using standard search engines with the use of publicly available sequences (IMGT, Uniprot, GenBank, RefSeq, for example).In FIG. 5C, the different regions are assembled into unique sequences using a combinatorial approach. Those sequences are assembled into a database (CA-FASTA DBS) and are then used in conjunction with a search engine to search the raw LC-MS data. A list of peptides and possible candidates is generated using standard proteomics search engines (Mascot, Sequest, Novor Cloud, etc.). The list of potential candidate sequences is reduced using an algorithm that mainly weights peptide overlap quality. This algorithm reduces the number of potential candidates based on the quality of the different peptide overlap and can be used in two different manners:1) High throughput: a minimum overlap score (MO score) value is evaluated for all possible sequences. Sequences having a MO score of zero are discarded.2) A more detailed analysis based on the overall quality of the overlapping profile allows ranking of possible candidates upon visual inspection.The present disclosure also provides a method of assembling complete or partial antibody protein sequences from a mix of peptides and / or a list of defined regions (e.g., multiple CDR1 , 2, 3, multiple FR1 , 2, 3 and JC regions) obtained from a plurality of unique antibodies (i.e., a polyclonal mixture), the process comprising any single or a combination of the following approaches: utilizing a combination of either digestion under non-reduced conditions or crosslink to gather evidence of distant region pairing (2 peptides linked by non-peptidyl bond) analysed by LC-MS and peptide match using a search engine that can process LC-MS data from crosslink experiments (either naturally crosslink by disulfide bridge or chemically crosslink); separating intact or reduced antibodies but kept as intact chains. Separating those chains into fractions may be done by gel or chromatography, the fractions are then digested with a single or multiple proteases and analyzed by LC-MS, the peptides identified and quantified, and the assembly of close and distant regions based on the profile similarity across the different fractions to complete the entire chain assembly; and generating peptides spanning distant regions (e.g., a peptide that includes a CDR1 , FR2, and CDR2) using specific proteases or chemical cleavage methods. The sequencing of those peptides is either performed de novo or using a search proteomics engine in combination with a FASTA database.The present disclosure also provides a method for assembling complete antibody protein sequences from a list of multiple mixes of peptides and / or defined regions (e.g., multiple CDR1 , 2, 3, multiple FR1 , 2, 3 and JC region) comprising a plurality of unique antibodies, wherein the method comprises:(1) utilizing a combination of either digestion under non-reduced conditions or crosslink to gather evidence of distant region pairing by LC-MS and peptide match using a search engine that can process crosslink data (either naturally crosslink by disulfide bridge or chemically crosslink);(2) separating intact or reduced antibodies but kept as intact chains, or intact domains (Fab / Fab2), and separating those chains by gel or chromatography into fractions, which are then digested with single or multiple proteases and analyzed by LC-MS, the peptides identified and quantified, and the assembly of close and distant regions based on the profile similarity across the different fractions to complete the entire chain assembly; and(3) using specific proteases of chemical cleavage methods to generate long enough peptides to span distant regions (e.g., a peptide that includes a CDR1 , FR2, and CDR2). Thesequencing of those peptides is either performed de novo or using a search engine in combination with a FASTA database.The present disclosure also provides a method for selecting potential sequence candidates based on the quality of the overlap of the different shorter peptide sequences when merged, they give a longer sequence. Those shorter peptides are the result of the digest performed with various proteases in parallel on a polyclonal antibody mixture. The method comprises: analyzing the peptides using LC-MS, and a FASTA database. The FASTA database is generated using either a combinatorial or a more restricted combinatorial approach to generate antibody sequences from a list of a plurality of FR1 , 2, 3 and CDR1 , 2, 3; generating peptide sequence and protein sequence candidates associated with confidence values from the list of identified peptides which were found using a standard proteomics search engine (Novor Cloud, Sequest, Mascot, and Maxquant); vectorizing the identified short peptide sequences by assigning a strength factor of either 0, 1 , 2, or 3 to each amino acid composing the short peptide sequence (a value of 0 to both the C and N-terminal, then a value of 1 for the C and N-terminal penultimate amino acid than all other amino acids position in the peptides sequence have a value of 3); calculating the quality of the overlap per each amino acid within a candidate sequence by summing all the numerical values resulting from the vectorization of all possible peptides composing the sequence candidate, wherein the 3 to 5 amino acids at the N- and C-terminal ends of each sequence candidate are discarded (i.e., to reduce the importance of the edge-effect); wherein the ranking of all sequence candidates is performed based on the highest minimum overlap score.Protein candidates are discarded if they contain “0” values within their sequences (typical of lack of overlap), and if multiple candidate sequences are kept, an overlap score profiling is then plotted in function of the amino acid position, and the overlapping quality evaluated manually to select a possible candidate or various candidates.The present disclosure also provides a method for the selection of potential sequence candidates based on the quality of the overlap of the different peptide sequences fitting a given sequence resulting from a digest performed with various proteases in parallel on a polyclonal mixture, wherein the method comprises: analyzing the peptides using LC-MS, and generating a FASTA database using either a combinatorial or a targeted approach to generate antibody sequences from a list of a plurality of FR1 , 2, 3 and CDR1 , 2, 3, JC region, and optionally including the hinge region; generating peptide sequences and protein sequence candidates with confidence values from the list of peptides using a standard proteomics search engine (e.g., Novor cloud, Sequest, Mascot, and Maxquant);vectorizing the identified peptide sequences by assigning a strength factor of either 0, 1 , 2, or 3 to each amino acid composing the peptide sequence (a value of 0 to the C and N-terminal, then a value of 1 for the C and N-terminal penultimate amino acid than all other amino acids position in the peptides sequence have a value of 3); calculating the quality of the overlap per each amino acid within a candidate sequence by summing all the numerical values resulting from the vectorization of all possible peptides composing the sequence candidate; wherein the 3 to 10 amino acids at the N- and C-terminal ends of each sequence candidate are discarded; wherein the ranking of all sequence candidates is performed based on the highest minimum overlap score, with protein candidates discarded if they contain 0 values within their sequences, and if multiple candidates kept an overlap score profiling plotted, and the overlapping quality evaluated manually to select a possible candidate or various candidates.The antibodies or antibody chains in the sample may be subjected to proteolytic cleavage with a suitable agent to generate the digested peptides. Agents to cleave proteins include chemical agents such cyanogen bromide (CNBr) that cleaves at methionine (Met) residues; BNPS-skatole that cleaves at tryptophan (Trp) residues; formic acid that cleaves at aspartic acid- proline (Asp-Pro) peptide bonds; hydroxylamine that cleaves at asparagine-glycine (Asn-Gly) peptide bonds, and 2-nitro-5-thiocyanobenzoic acid (NTCB) that cleaves at cysteine (Cys) residues, as well as enzymes (e.g., proteases). In an embodiment, the antibodies or antibody chains in the sample is subjected to proteolytic cleavage with any suitable enzyme (e.g., protease) or combinations thereof. Enzymes that may be used to perform the proteolytic cleavage include trypsin, pepsin, thermolysin, chymotrypsin, AspN, LysargiNase, LysC, LysN, GluC, ArgC, Pro / Ala protease, Sap9, KEX2, or any combinations thereof. The method may also include a step of modifying the cysteines to convert them into a lysine analog, for example using electrophilic ethylamine, for example 2-Bromoethylamine hydrobromide (BEA) that alkylates cysteine residues (to aminoethylcysteine residues), which can then generate cleavage sites for proteases such as Lys-C and trypsin for example. The digestion may also be performed with a combination of chemical agent(s) and protease(s). The digestion can be allowed to go to completion so the protein is cleaved at all bonds that the digestion reagent is capable of cleaving; or the digest conditions can be adjusted so that fragmentation does not go to completion deliberately, to produce larger fragments that may be particularly helpful in determining the sequence of distant regions in the antibody chain; or digest conditions may be adjusted so the protein is partially digested into domains. The conditions that may be varied to modulate digestion level include duration, temperature, pressure, pH, absence or presence of protein denaturing reagent, the specific protein denaturant (e.g., urea, guanidine HCI, detergent, acid-cleavable detergent, methanol, acetonitrile, other organic solvents), the concentration of denaturant, the amount orconcentration of cleavage reagent or its weight ratio relative to the protein to be digested, among other things.The methods described herein may be used to identify one or more antibodies or antibody chains from an antibody mixture such as a polyclonal antibody mixture or a mixture of monoclonal antibodies. The mixture may be derived from a biological sample (e.g., blood sample) from a subject, for example a subject who has been vaccinated or infected by a microorganism (e.g., a virus). The polyclonal antibody mixture may be collected either from an animal(s) or from tissue culture supernatants of B cells. In various embodiments of the non-limiting methods of the disclosure, the polyclonal antibody mixture may have, for example, at least two different immunoglobulins within the mixture, or at least three, or at least five, or at least ten, or at least twenty, or at least fifty, or at least one hundred or at least five hundred different immunoglobulins within the mixture.The antibody mixture collected, either from an animal(s) or from tissue culture supernatants of B cells, can be subjected to a protein A or protein G Sepharose™ column, which can separate antibodies from other blood sera proteins, for example. Additionally, the collected polyclonal antibodies may be subjected to antigen affinity purification to enrich for antibodies with high specific activity. A purification step, such as antigen affinity purification, may be performed to reduce the complexity of a polyclonal mixture and ultimately reduce the number of potential false positives. The collected polyclonal antibodies may be concentrated or buffer exchanged or both, either before or after purification.The antibody that may be identified by the methods described herein can either be conventional IgG (paired heavy and light chains sequence) or non-canonical antibody (heavy chain antibody from camelid such as Alpaca, Llama, camel), as well as IgY from avian species (e.g., chicken).In an embodiment, the method further comprises assessing the binding of a recombinant antibody or antibody fragment generated from the sequence information obtained by the method described herein to its target antigen. This step may be used to confirm that the putative antibody identified by the method forms a functional antibody or antibody fragment, or in the case where more than one putative / candidate antibodies are identified, this step permits to confirm which of the putative / candidate antibodies is a functional antibody or antibody fragment able to bind to the target antigen. This may be done, for example, by recombinantly expressing an antibody or antibody fragment corresponding to a putative antibody identified, and assessing the binding of the recombinant antibody or antibody fragment to its target antigen. This may be achieved by introducing nucleic acids encoding the heavy and light chains of the putative antibody into a suitable expression system, such as CHO cells, HEK293 cells, yeast cells, or E. coli cells, culturing the cells into conditions suitable for the production of the antibody or antibody fragment, and assessing the binding of the antibody or antibody fragment to the antigen, for example byimmunoassay (e.g., ELISA) or by surface plasmon resonance (SPR). In an embodiment, the binding of the antibody or antibody fragment to its target antigen may be assessed by expressing the antibody or antibody fragment at the surface of a phage and assessing the binding of the phages to the antigen (phage panning).EXAMPLESThe present disclosure is illustrated in further detail by the following non-limiting examples.Example 1 : Sequencing of two antibodies in an artificial mixtureAn embodiment of the methods of the present disclosure is demonstrated by Example 1 , wherein two commercially available antibodies with known sequences were mixed, separated under non-reducing conditions, fragmented using different proteases, sequenced using LC-MS and MS / MS, and assembled to generate an accurate, complete antibody sequence.In this example, chromatography separates the two standard antibodies under non-reducing conditions, followed by protease digestion of the fraction and chain assembly based on peptide identification, peak intensity, and intensity correlation between the different peptides.The correct assembly of two mice lgG2a from an artificial mixture is shown using Hydrophobic Interaction Chromatography (HIC) to separate the antibodies mixture into fractions (FIG. 6).Chromatography was used to separate two standard mixed antibodies in non-reducing conditions, followed by protease digestion of the fraction and chain assembly based on peptide identification, peak intensity, and intensity correlation between the different peptides. Through hydrophobic interaction chromatography (HIC), 50 pg of each intact antibody was mixed and separated, resulting in two prominent peaks. Subsequent fractions were reduced with dithiothreitol (DTT), alkylated with iodoacetamide (IAA), and digested using chymotrypsin and pepsin proteases.The digested fractions were then analyzed using liquid chromatography-mass spectrometry (LC-MS). Identification of peptides per fraction was performed with Novor search engine, which contains an internal database of in-house antibody sequences. Peaks associated with the selected sequences representing any heavy complementarity determining regions (CDRs) for both mixed antibodies were then extracted and evaluated for each fraction. Intensity data demonstrated that the CDRs clustered together in a given fraction based on their origin (Table 1). Some peaks labeled as “contamination” were identified, resulting from sample carry-over, and are around 3-5% of the main signal and thus negligible. Pair scatter plots were produced from the inverse hyperbolic sine function (arcsinh) transformed data, and correlations were extracted between the different CDRs (Table 2).Table 1Table 2This example shows that by using a mix of two monoclonal antibodies, the different CDRs regions for each of the two antibodies can be easily assembled correctly using a separation method, in this case, hydrophobic interaction chromatography.Example 2: Assembling a Heavy chain-based antibody from a natural mixture (Alpaca) -Generating a pool of individual regions.MethodsA method for assembling a heavy chain-based antibody from a natural mixture (Alpaca) is provided herein. A commercial polyclonal antibody mixture from Alpaca was used. It contains canonical IgG (lgG1a, b) and non-canonical heavy chain only IgG (lgG2b,c). Although the Alpaca antibody naming is somewhat contradictory, the nomenclature from IMGT was chosen (https: / / www.imgt.org / ). The antibody was labelled PD106 (internal reference) and was prepared for a 5-enzyme digest. The sample was diluted with HPLC grade water and then reduced with DTT at a final concentration of 30 mM with heating at 95°C for 15 minutes, allowed to cool to room temperature, then alkylated with lodoacetamide (IAA) at a final concentration of 50 mM for 30 minutes in the dark at room temperature. The sample was precipitated by adding three times the sample volume of acetone and then incubated at -20°C for 1 hour. The sample was pelleted in a centrifuge at 23,000 x g for 10 minutes at 4°C; acetone was carefully decanted to avoid disrupting the pellet, then the pellet was dried under a low-pressure centrifuge (SpeedVac™). The pellet was resuspended in 10 pL 4M urea at 37°C for 10 minutes to ensure complete dissolution, then diluted up to 100 pL with HPLC grade water and separated 20 pL across five tubes for the five digestions. Pepsin, Trypsin, Chymotrypsin, AspN, and LysC enzymes are used to perform the digestion. The samples were reconstituted in 40 pL 0.1% FA, and 5 pg were loaded on Evotips. Samples were analyzed on a Thermo Orbitrap™ Exploris 240 using a 44-minute method, and injection of selected samples (AspN, LysC, and trypsin) was also performed in Middle-down proteomics mode. Dried peptides were resuspended in 0.1% FA at one pg / pl, and five pg was loaded into an HPLC vial. Two HPLC column systems (Thermo Scientific PepMap 75pm x15cm, 3pm, 100A column and a C18 column from WATERS, nanoEASE™ M / Z peptide BEH C18, 300A, 1.7pm, 300pm x 100mm) were used. A mass spectrometric analysis is performed on an Orbitrap™ Eclipse (Thermo Scientific), with a survey scan done at 60,000 resolutions. Dynamic exclusion was enabled for 15s duration, and HCD with a relative collision energy of 35 was applied on isolated precursors from charge states 2-11. Product ions are detected in the Orbitrap™ at 30,000 resolutions with a maximum injection time of 150ms.SequencingSeveral regions were found de novo using Novor Cloud search engine and searched against germline. They are shown in Table 3; the separation into different regions was based on the information found in IMGT. However, several experimentally found sequences were not in the IMGT database. Most of the FR1 ends up with "CAAS" with a serine, “S”. At the same time, CDR1 starts with a glycine, G. The division between regions could sometimes be ambiguous and was defined by analyzing overlapping peptides. For FR2, a dominating "QAP" around 4-6 aa is found from the FR2 N-term. Framework 3, FR3 is easily defined with a cysteine in the third position of the C-terminal end. The short de novo regions presented in Table 3 can be generated using any de novo software such as Novor search engine or Peak (Bioinformatics Solution).Table 3 a b c d672910I hJ C / hlnge for lgG2b,c1 hinge for I G 2b2 hinge for IgG 2cEVHLVESGGGLVQPGGSLRLSCVVS (SEQ ID N0:7); KVQLVESGGGLVQAGGSLRLSCAAS (SEQ ID N0:8); QVQLVESGGGLVQAGDSLRLSCAVS (SEQ ID N0:9); QVQLVESGGGLVQTGGSLRLSCAAS (SEQ ID NQ:10); QVQLVESGGGLVQTGGSLRLSCALS (SEQ ID N0:11); GFDFDDW (SEQ ID N0:12); GFRFSFY (SEQ ID N0:13); GFTFDDF (SEQ ID N0:14); GGTFNAY (SEQ ID N0:15); GVDSISDMS (SEQ ID N0:16); DMSWFRQAPGKEREGVSCI (SEQ ID N0:17); FMAWFRQAPEKEREFVTRI (SEQ ID N0:18); QMSWVRQAPGKGLEWLATI (SEQ ID N0:19); TAGWFRQAPGKAREFLASI (SEQ ID NQ:20); TIGWFRQAPGKEREPVSCI (SEQ ID N0:21 ); NNNGDS (SEQ ID NO:22); NWSGKF (SEQ ID NO:23);SKRDGL (SEQ ID NO:24); SRHDDM (SEQ ID NO:25); SRRDGR (SEQ ID NO:26);INYADSVKGRFTIGRDTAKNTVYLQMNSLKPEDSAVYYCAA (SEQ ID NO:27);PRYPDSAEGRFTISRDNAKNTLYLQMNSLKPEDTAVYYCAK (SEQ ID NO:28);TRYGDSVKGRFTISRDNAKEMAFLQMNSLKPEDTAIYYCVA (SEQ ID NO:29);TYYADSVKGRFTFSSDNAKRTVYLQMNSLKPEDTAVYYCAA (SEQ ID NQ:30);TYYTDSVKDRFTISRDNANKVVFLQMNGLKPEDTAVYYCAA (SEQ ID NO:31 );DEPPYRCSDYWEPWREY (SEQ ID NO:32); DEPPYRCSGGWEPWREY (SEQ ID NO:33);DEPPYRCSSSWDPWREY (SEQ ID NO:34); DEPPYRCSSSWTPWREY (SEQ ID NO:35);GSTWGVSGRRVPDYDY (SEQ ID NO:36); GTTWGVSGRRVPDYDY (SEQ ID NO:37);PKFPRLDQWVTWDELDY (SEQ ID NO:38); RILCPLDWSPRHEYTEEVVGS (SEQ ID NO:39);RILCPMDWNSRREYTEEVVGS (SEQ ID NQ:40); RLQGLSAEAEEYDF (SEQ ID NO:41); WGQGAQVTVSS (SEQ ID NO:42); WGQGTQVTVSS (SEQ ID NO:43); EPKTPKPQPQPQPQPQPNPTTESKCPKCP (SEQ ID NO:44); AHHSEDPSSKCPKCP (SEQ ID NO:45).A combinatorial approach was used to generate the list of potential Fab sequences. All sequences have a similar structure and are based on the information extracted from Table 3. The naming strategy is the following: Seq-a-b-c-d-e-f-g, where the “a” is a vector “FR1”; thus, if a = 1 , then the FR1 is (information taken from Table 3 the sequence is the first entry of the FR1 vector “EVHLVESGGGLVQPGGSLRLSCWS” (SEQ ID NO:7)). This corresponds to the first entry in the “vector” FR1 .The sequence Seq-1 -2-3-1 -2-7-1 is, therefore, the first entry in vector FR1 , the second entry in vector CDR1 , then the third entry FR2, the first entry in vector CDR2, the second entry in FR3, the seventh entry in the CDR3 vector, and finally the first entry in FR4 region. FIG. 7 illustrates the principle and the link between sequence naming and how to assemble them. This naming convention allows linking the sequence naming to a given sequence and allows the visualization of a trend in sequences easily (simply based on the naming strategy used), thus facilitating the analysis. Although several CDR3 were identified, only 10 sequences are depicted in Table 3. Moreover, a total of 10 working heavy chain-based antibodies were identified. This example shows how to construct only two candidate sequences. Those two full constructs are designated "R1" and "R2". The sequence R1 will be based on CDR3[7] (PKFPRLDQWVTWDELDY, SEQ ID NO:38) and R2 based on CDR3

[0010] (RLQGLSAEAEEYDF, SEQ ID NO:41). It was decided to build antibodies from their more unique aspects, which is the CDR3.A global FASTA database from those seven regions, i.e., FR1 , CDR1 , FR2, CDR2, FR3, CDR3, and FR4, was first generated. The inclusion of the hinge region is only done later in order to reduce the number of combinations to explore. The overall size of the database can be challenging to handle; in this present case, it is a total of at least: 5x5x5x5x5x10x2 possibilities, which is 62,500 sequences. If the analysis is limited per CDR3 region, the database size to generate is divided by 10 (6,250 sequences). For each CDR3, a combinatorial database was built with all the possibilities (germline usage was not considered for this particular example, but this is a possibility that could reduce the number of possible combinations to generate). Thus, for the two sequences, R1 and R2, we then have: For R1 : Seq-a-b-c-d-e-7-g and R2: Seq-a-b-c-d-e-10- g Where a, b, c, d, e, and g can take any values within their respective "region vector" (i.e., "a" from FR1 will take sequence value from 1 to 5, same for b, c, d, and e, one only CDR3 at a time and 2 J / FR4 regions).A total of 5x5x5x5x5x1x2 = 6,250 sequences were generated per specific CDR3 (either ranked "7" or "10" within the vector CDR3 described herein). Those sequences were generated and assembled using the program R (i386 3.6.3) Stringr and Sequinr package and in-house script to generate the FASTA file and analysis. Each FASTA dataset was then generated and usedagainst the initial experimental dataset (a mix of several LC-MS runs done with trypsin, pepsin, chymotrypsin, AspN, and LysC). The data were searched using the online tool Novor Cloud (https: / / app.novor.doud / home), although any database search engine could have been used, such as Mascot, Sequest, and Maxquant to name a few. One main advantage of using Novor Cloud is its ability to handle most of the combination of different protease digestions which is advantageous for the final assembly for this particular application. The search criterion on Novor Cloud were as follow:FDR set at 1%, MS tolerance at 5ppm, and MSMS tolerance set at 100ppm, all modifications set as variable Pyro-Glu (E), Pyro-Glu (Q), Carbamidomethyl (C), and Deamidated (NQ)).One major issue with using a combinatorial FASTA database construct based on the most intense / abundant peptides found in a sample and searching those same or similar experimental datasets is that the obtained result favors any protein construct with the highest number of experimental evidence, although they can be false. To address this issue, an approach where protein candidates were ranked on the quality of peptide overlap is proposed, as illustrated in FIG. 8, where the same protein is digested with different specific proteases; thus, peptide overlap should be obtained.Here, in the case of A, the hypothetical protein has more experimental evidence (more identified peptides) than in case B using conventional proteomics identification software such as Novor cloud, Mascot, and Sequest. Directly using any of those search engines, case A will rank betterthan case B. A different approach, which favored the quality of the different peptide overlaps instead, was used in this study. Consequently, the identified peptide sequence was vectorized by assigning a "strength" factor of either 0, 1 , 2, or 3 to each amino acid composing the peptide sequence (a value of 0 to the C and N-terminal, then a value of 1 for the C and N-terminal penultimate amino acid than all other amino acid positions within the peptides sequence have a value of 3). The peptide vectors were projected into the correct position within a sequence candidate, then the overlap quality was calculated per each amino acid within a sequence by summing all the numerical values resulting from the vectorization for that particular amino acid position. With such an approach, one could look at the value of "0" within a sequence, which shows a poor overlap for case A. Protein match using an overlap calculation allows the ranking of protein matches based on the quality of the peptide overlaps instead of total peptide abundance. It does not favor hits based on a high number of peptides. In that example from FIG. 8, case B will have a higher "rank" due to a better overlap quality of the peptide, even though fewer peptides were found for that protein.Thus, under standard proteomics conditions, "case A" is shown as a top hit with any standard proteomics search engine; however, using the proposed approach, an overlap score, case B will rank higher. The choice of the peptide Vector containing values of 0, 1 , 2, and 3 maximum is based on a basic empirical observation; a maximum of "3" was chosen to avoidadding too much weight to a longer peptide which often contains less sequence information than a shorter one. The total overlap is calculated by projecting the total score per amino acid position. This strategy was used in two different ways possible:Approach 1: Filtering off protein candidates by minimum score close to zero (proteins having gaps in their sequence coverage were then eliminated). This approach allows to discard rapidly large number of protein candidates in a high throughput manner.Approach 2: Looking at the quality of sequence or region coverage. Proteins can then be ranked based on the overlapping quality, not the total peptide match. In this example, case A will rank higher from a total peptide identified with any of the conventional search engines. However, case B has a higher overlap score than case A based on the minimum overlap. Experimentally, amino acids at the protein extremities were discarded, as they will have a value of "0" (edge effect). The evaluation of the minimum overlap is sample dependent, and the parameters used are described in all the different examples where this overlap quality strategy is used. Approach 1 was used as a high throughput first-round approach to eliminate non-interesting sequences (to discard hundreds to thousands of sequences). Approach 2 was used on a few selected potentially suitable candidates. The following section illustrates how two heavy chain-based antibodies from a natural mixture of alpaca antibodies were sequenced.After generating the experimental dataset on the natural polyclonal mixture with several proteases (chymotrypsin, pepsin, AspN, trypsin, Lys-C), this same experimental dataset was used to search against the two different FASTA files: R1: seq-a-b-c-d-e-7-g and R2: seq-a-b-c-d- e-10-g.The initial search for CDR3[71, R1 and CDR3[101, R2For the CDR3[7] (referred to as R1), 1562 combinatorically constructed proteins were identified from a database of 6,250 sequences using Novor cloud. The edge effect was encountered as the protein extremity had a "0" value, as all peptides at the N-term and C-term had a "0". The analysis did not extend over the J region, which reduced the confidence of the measured overlap for the FR4. Due to the edge effect, anything from amino acid positions 1 to 4 (N-term) and anything after position 99 (referred to as "high") or after position 93 (referred to as "low") was removed from the overlap calculation. These limits were defined arbitrarily to identify the minimum within the variable part of the protein sequence without considering the extremity of the sequences.From the 1562 sequences kept by Novor cloud, only 66 combinatorically constructed proteins passed the criterion of not having within the arbitrarily defined boundary any null value. The top two candidates, having the best quality overlap, were ranked 57th and 844th by Novor cloud regarding total peptides found. They correspond to Seq-1-2-3-1-2-7-1 and Seq- 1-2-3- 1-2- 7-2, with a minimum overlap (MO) score of 6 in both cases. The list of the 66 combinatoricallyconstructed proteins is shown in Table 4. Interestingly, those top 2 candidates have the same FR1 ,CDR1 ,FR2,CDR2,FR3 and obviously same CDR3, and only differ from each other from the FR4 region.Table 4For the CDR3

[0010] (referred to as R2), a total of 571 proteins were identified. Again, to avoid the edge effect, amino acids in positions 1 to 4 and from 107 and above were removed as part of the minimum overlap calculation. From those 571 sequences, only nine selected sequences did not show null gaps within their sequence. All of the selected forms were dominated by FR4 "2". The sequence having the best MO score was ranked 361stby Novor cloud and has the sequence Seq-3-4-4-2-3-10-2, with a MO score of 5. The list of the nine sequences is shown in Table 5. To confirm distant region pairing, a complementary analysis was performed.Table 5For sample digestion under non-reduction and disulfide bridge pairing analysis, 20 pg of PD106 in duplicate was dried using low-pressure centrifugation (Speedvac™). Both samples were then reconstituted in 10 pL of 8M urea and one of the duplicates were treated with NEM at a final concentration of 2 mM before being incubated with shaking at 37°C for 2 hours. Subsequently, both samples were diluted up to 40 pL with 50 mM ammonium bicarbonate, pH 8, to achieve a final urea concentration of 2 M to allow for efficient digestion by Trypsin (Promega) at a 1 :20 enzyme-to-protein ratio. The samples were then incubated overnight at 37°C. The following day, the digested samples were diluted up to 80 pL with 50 mM ammonium bicarbonate, pH 8, to reach a final urea concentration of 1 M, thus accommodating the urea tolerance of GluC (Promega). GluC was added at a 1 :20 enzyme-to-protein ratio and then incubated with shaking for 4 hours. After drying the samples under low pressure (SpeedVac™), the samples were reconstituted in 40 pL 0.1% FA, and 5 pg were loaded on Evotips according to the manufacturer’s instructions. The samples were then run on a Thermo Orbitrap™ Exploris 240 using a 44-minute gradient method. The identification of the S-S bridge containing peptides was performed using pLink v2. The parameters used for software identification were set as follows: Flow Type: Disulfide bond (HCD-SS); Enzyme: LysC-AspN, Try, orTry-GluC; Peptide mass: 300-9000; Peptide length: 3-90; Fixed modification: Gln->pyro-Glu; and Variable modifications: “Nethylmaleimide[C],” “Oxidation[M],” “Deamidated[N],” “Deamidated[Q],” and “Acetyl[ProteinN-term].” The protein database contained the sequences of interest for the pAb or mAb mix of interest and the corresponding reversed sequences. Table 6 illustrates some identified peptides, with the peptide assignment restricted to a specific region (FR1 , CDR1 , FR2, etc.). In some cases, the information obtained from the peptide was not enough to specifically identify a single given region, so a range was proposed. In other cases, the distant pairing of specific regions, the obtained information can be precise such as in peptide #2 and #4 where a unique FR1 and CDR1 is paired to a unique FR3 and CDR3 from Table 3.Table 6DTAVYYCAK (SEQ ID NO:46); LSCVVSGFR (SEQ ID NO:47); DTAVYYCAKPK (SEQ ID NO:48); NTLYLQMNSLKPEDTAVYYCAK (SEQ ID NO:49); NTLYLQMNSLKPEDTAVYYCAKPK (SEQ ID NO:50); MAFLQMNSLKPEDTALYYCVAR (SEQ ID N0:51 ); LSCAVSGGTFNAYTAGWFR (SEQ ID NO:52); VVFLQMNGLKPEDTAVYYCAADEPPYR (SEQ ID NO:53); LSCAASGFDFDDWTLGWFR (SEQ ID NO:54); NTVYLQMNSLKPEDSAVYYCAAGTTWGVSGR (SEQ ID NO:55); LSCALSGVDSLSDMSFMAWFR (SEQ ID NO:56); LLCPMDWNSR (SEQ ID NO:57); EGVSCLSR (SEQ ID NO:58); EREGVSCLSR (SEQ ID NO:59); LLCPMDWNSRR (SEQ ID NQ:60); LLCPLDWSPR (SEQ ID N0:61);Table 6’s disulfide bridge information, when combined with the prior search, confirms the two sequences R1 (seq-1 -2-3-1 -2-7-g) and R2 (seq-3-4-4-2-3-10-2). The SS search was not able to fully confirm CDR3

[0010] , but presented two other possibilities, CDR3[8] and CDR3[9],Regarding R1 , a distinct decision must be made on the suitable J segment for two reasons. Firstly, neither of the likely J peptides possess a “protease friendly cutting site”, which means that none of the employed proteases will generate a peptide that will contain a unique element of one of the two possible J regions, thus, the J segment is poorly covered. To guarantee good coverage of the J region, extending the database past the J region into the hinge region is important. Secondly, this polyclonal sample mix includes a combination of lgG1a,b (standard antibody with heavy and light chain) and lgG2b,c (non-canonical, just the heavy chain type, which is the focus of the present study). The canonical antibody sequence transitions from the J region to the C region commencing with the CH1 domain; thus, the generated peptides are from J to CH1 and have the following structure ... “TVSSASTK” ... the lysine, K is a trypsic and a lysC site, making this peptide highly abundant for lgG1a,b. For the non-canonical antibody (the one that is targeted in this study as Alpaca antibodies do not have a light chain), the transition is from the J region to the hinge region. Two different peptides for Alpaca define the hinge (see Table 3): EPKTPKPQPQPQPQPQPNPTTESKCPKCP (SEQ ID NO:44) as “FR5(1)” and AHHSEDPSSKCPKCP (SEQ ID NO:45) as “FR5(2)”, which were added to a small database of possible sequences: For R1 : Seq-1 -2-3- 1-2-7-g-h For R2: Seq-3-4-4-2-3-(8-10)-g-h where “g” is the J / FR4 (with two possible values), and “h” is “FR5” (as the hinge peptide) with two potential values too. All the rest of the sequence was already confirmed. It is important to keep in mind that the sample has not been fractionated and encompasses a combination of all digested peptidesfrom the canonical and non-canonical antibodies, which could interfere with the correct assignment. Adding the hinge regions as a strategy will assist in verifying the proper J region.R1 final sequencing: Seq-1-2-3-1-2-7-q-hIn FIG. 9A, overlap scores of four different sequences (seq-1-2-3-1-2-7-g-h, where g and h can have two different values) are plotted against the peptide raw experimental dataset 1 . There is no significant difference between the four sequences, from FR1 to FR3. However, a difference can be seen when comparing the transition from FR3 to CDR3 (highlighted by the two arrows), depending on the choice of the J / C and hinge region peptide. The best sequence overlap is generated with g=1 and h=2. Under those conditions, the peptide “LDQWVTWDELDYWGQGAQVTVSSAHHSE” (SEQ ID NO:62) at 1082.1584 amu with an intensity of 7.3e9 has been detected. This peptide covers the CDR3[7], FR4[1], and the hinge region FR5[2] well. Although the peptide is non-tryptic, the actual natural sample shows some degradation of the VHH protein, with a band around 15kDa (FIG. 9B). This corresponds to a fragment of VHH that binds to the antigen (rabbit IgG) which would end at the C-terminal end with ...VSSAHHSE; thus, the R1 sequence will most likely be Seq-1-2-3-1-2-7-1 . The small 15kDa sequence found in the sample was investigated in the next section.Middle Down sequence confirmation for R1 :The LC-MS dataset generated for the study was searched against a database containing all possibilities of CDR3[7] for R1 using the same search criteria as described above. Any long peptides that cover at least 3 regions, thus satisfying the criterion of pairing distant regions, were looked for. The results are reported in Table 7. For R1 , Table 7 reports 4 different peptides each covering 3 to 4 regions, including FR1 , CDR1 , FR2, CDR2 and FR3. Through this “middle down” approach, it is possible to associate a CDR1 to a CDR2 peptide for example.Table 7EVHLVESGGGLVQPGGSLRLSC(Cam)VVSGFRFSFYQMSWVRQAPGK (SEQ ID NO:63);YQMSWVRQAPGKGLEWLATINNNGDSPRYPDSAEGRF (SEQ ID NO:64);ATINNNGDSPRYPDSAEGRFTISRDNAKNTLY (SEQ ID NO:65);GLEWLATINNNGDSPRYPDSAEGRFTISRDNAK (SEQ ID NO: 66)Experimental description of the natural Alpaca fraction binding to the antigen (i.e., rabbit IgG).The aim of this experiment is to attach rabbit IgG, rabbit IgG Fc, and rabbit IgG F(ab)2to agarose-based beads and then use immunoprecipitation to assess how the commercial Alpaca polyclonal antibody binds to its antigen, a rabbit IgG. For the antigen fragment production, 100 pl of 10 mg / pl Rabbit IgG Sigma l8140-10mg lot SLBK40780 was mixed with 100 pl PBS and 20 pl of IdeS (50u / pl, Promega) and incubated for one hour at 37°C. To generate the F(ab)2and Fc fragments, 0.3 mL of Genscript™ Protein A resin FF (cat no L00464-5) was used on a Sigma (C2728) fritted column. The resin was conditioned twice with 0.75 mL PBS; then the IdeS digested samples were loaded. The sample was re-loaded three times, and the flow-through (FT) was kept (corresponding to the Fab2fraction). The column was then washed with 0.3 mL PBS, and the wash was combined with the FT. Three additional washes of 0.5 mL PBS were subsequently performed and discarded. The Fc was eluted from the A-protein beads using 0.1 mL glycine pH 2.7 in a neutralized buffer of 0.124 mLtris 1 M. Buffer exchange was carried out using an Amicon™ filter 3kDa (Sigma). A single Pierce tube Pierce NHS-Activated was split into three tubes, Mobicol™ fitted with a filter of 10 pm pore size (Boca scientific). The three forms of antigens were coupled to the NHS beads:1) Total rabbit IgG,2) F(ab)2rabbit,3) Fc rabbit.A second Pierce NHS-Activated Agarose Spin Column was used as a negative control and neutralized with tris 1M only. The four tubes were incubated with their respective antigen overnight, then washed with water and neutralized with 20 pl tris 1M. 25 pg of the alpaca antirabbit polyclonal antibody was added to each of the four columns, incubated for one hour, then eluted and washed with 50 pl of water. The binding fraction was eluted using SDS-PAGE loading buffer and heat 95°C for 5 minutes. The gel was run (see FIG. 9B). Each gel band was trypsinized using a standard protocol (in gel digest) and run on a Thermo Orbitrap™ Exploris 240 with a 44min Evosep™ gradient. The results showed that a fraction of the Alpaca antibody bound to the beads (lane Tris E), and two prominent bands were found at 150 kDa (labeled 1) and 15 kDa (labeled 7) for most of the other three samples (total rabbit antigen, rabbit Fab2, and rabbit Fc). The band of 15 kDa fit the variable domain of the VHH and the tryptic digest showed the same peptide LDQWVTWDELDYWGQGAQVTVSSAHHSE (SEQ ID NO:62) mentioned above, which suggest that this domain seems to cleave from the chain through a yet unknown mechanism. This peptide resulting from trypsin digest was semi-tryptic with a mass of 1082.1598 amu as a 3+ and was a peptide from cdr3 (LDQWVTWDELDY, SEQ ID NO:67), J region (WGQGAQVTVSS, SEQ ID NO:42), and the hinge from lgG2c (AHHSE, SEQ ID NO:68); thus, it contained no CH1 domain, which is typical of the non-canonical antibody from Alpaca.Final sequencing for R2: Seq-3-4-4-2-3-(8,9,10)-q-hFrom the twelve possibilities (3x2x2), Novor cloud identified ten of the most plausible sequences, as demonstrated in FIG. 10A. This heatmap shows the score overlap against the amino acid position, where white indicates a low score overlap and black indicates a high score overlap. The two arrows at the top of the heatmap point to the N-terminal and the C-terminal side, respectively, of the CDR3 region, showing that CDR3

[0010] (RLQGLSAEAEEYDF, SEQ ID NO:41) is the most probable among the 3 selected CDR3, with four candidates containing this CDR3 sequence (2nd, 4th, 6thand 10thposition). FIG. 10B displays a representation of these four candidates which have this particular CDR3

[0010] , with two different FR4 and two different hinge peptides. G02 and G04 are the most likely sequences as they both show higher than zero value for the overlap score in the CDR3 region. There is evidence of both peptides in the sample (Seq- 3-4-4-2-3-10-2-1 and Seq-3-4-4-2-3-10-2-2, which only differ in their hinge regions, thus different isotypes). Therefore, the most probable sequence for R2 is both hinge possibilities combined: Seq-3-4-4-2-3-10-2.Testing recombinant sequences R1 and R2 against its antigen:Recombinant forms of human IgG 1 heavy chain Fc and alpaca variable region PD106_R1 (Seq-1 -2-3-1 -2-7-1) and PD106_R2 (Seq-3-4-4-2-3-10-2) were created. The human Fc lgG1 domain CH2 was “DKTHTCPPAPELL...” (SEQ ID NO: 217), which can be seen in its entirety at IMGT (accession No. J00228). Affinity of the recombinants against rabbit IgG was evaluated using ELISA with anti-human HPR species as a secondary antibody. FIG. 11 B revealed PD106_R1 and R2 had strong affinity against the antigen with a 50% OD 450nm below 1 nM. Two negative controls were used: a human IgG isotype and a recombinant non-binder named PD106_R10.Example 3: Sequencing of a rabbit polyclonal antibody into a paired heavy and light chain (PD108)Heavy and light chains from a polyclonal antibody mixture obtained from two rabbits immunized with the short peptide sequence FPPSSEEL (SEQ ID NO:68) were sequenced. To pair the heavy and light chains, the method described in PCT / CA2022 / 051194 was used. 14 recombinant forms were generated and only five displayed a similar affinity to the original natural polyclonal antibody (designated PD108). The experiments focused on a single form called PD108_r1 , which combines a given Heavy chain (named Heavy 1) with a given Kappa chain (named Kappa 1). The objective of this example is to show how it is possible to de novo sequence individual Heavy and Light chains in a complex mixture of similar antibodies. A list of different regions (FRs and CDRs shown in Table 8) was generated from a comprehensive digest of the polyclonal mixture, combined with a de novo sequencing and database search using the common rabbit germline (IMGT). The right regions were then combined to generate a functional chain, which involved combinatorial database generation, scoring sequence overlap, and the use of non-reduced sample digest to reduce the number of combinations for pairing distant regions. This sample - PD108, a rabbit polyclonal antibody against FPPSSEEL was obtained from Vivitide.Table 8Heavy chain45€ a b e dLight chainFR1 CLR1 FR2 CLR2e f g910QSVEESGGRLVTPGGSLTLTCTVS (SEQ ID NO:69); QSVEESGGRLVTPGTPLTLTCTVS (SEQ ID NOTO); GFSLNNY (SEQ ID NO:71 ); GFSLNTY (SEQ ID NO:72); GFSLSSY (SEQ ID NO:73); GFSLSTY (SEQ ID NO:74); AMGWVRQAPEKGLEYIGII (SEQ ID NO:75); AMGWVRQAPGKGLEYIGII (SEQ ID NO:76); GVSWVRQAPGKGLEFI (SEQ ID NO:77); PMAWVRQAPGKGLEWIGWI (SEQ ID NO:78); PMGWVRQAPGKGLEWIGWI (SEQ ID NO:79); SMSWVRQAPGKGLEWIGII (SEQ ID NQ:80); ANSGY (SEQ ID NO:81 ); GFIRASGS (SEQ ID NO:82); GPISG (SEQ ID NO:83); GSSGN (SEQ ID NO:84); GTITG (SEQ ID NO:85); ARYANWAKGRFTISRTSTTVDLKMTSLTTEDTATYFCAR (SEQ ID NO:86); ARYASWAKGRFTISKTSTTVDLKMTSLTTEDTATYFCAR (SEQ ID NO:87); AWYASWVKGRFTISKTSTTVDLKITSPTTEDTATYFCTR (SEQ ID NO:88);RYYASWTKGRFTISKTSTTVDLKVTSPTTEDTATYFCAT (SEQ ID NO:89);TYYASWAKGRFTISKTSTTVDLKITSPTTEDTATYFCAR (SEQ ID NO:90); DGDSINGAVMDF (SEQ ID NO:91 ); DGDTTTGAVMDL (SEQ ID NO:92); DGNSAYNSGVNL (SEQ ID NO:93); NLNAVSATGHTFDP (SEQ ID NO:94); NLNVVSSTGHAFDP (SEQ ID NO:95); SLFGSGDVDNL (SEQ ID NO:96);WGPGTLVTVSSGQPK (SEQ ID NO:97); WGQGTLVTVSSGQPK (SEQ ID NO:98);WGRGTLVTVSSGQPK (SEQ ID NO:99); ADIVMTQTPASVEAAVGGTVTIKC (SEQ ID NQ: 100);ADVVMTQTPSPVSAAVGGTVSISC (SEQ ID NQ:101 ); ALVMTQTPASVEAAVGGTVTINC (SEQ ID NQ:102); AYDMTQTPASVEAAVGGTVTIKC (SEQ ID NQ:103); DVVMTQTPSSVEAAVGGTVTIKC (SEQ ID NQ:104); QAVVTQTPSSVSAAVGGTVTISC (SEQ ID NQ:105); QVLTQTASPVSAAVGGTVTINC (SEQ ID NQ:106); QVLTQTASPVSAAVGSTVTINC (SEQ ID NQ:107); QASQSISGSYLA (SEQ ID NQ:108); QASQSISNLLA (SEQ ID NQ:109); QASQSISNQLS (SEQ ID NQ:110); QASQSISSYLA (SEQ ID NO:111); QASQSVYNNDRLS (SEQ ID NO:112); QASQSVYNNNLA (SEQ ID NO:113);QSSKSVDKDNRLA (SEQ ID NO:114); QSSKSVYNHNRLS (SEQ ID NO:115); WFQQKPGQRPKLLIY (SEQ ID NO:116); WLQQKPGQPPKRLIY (SEQ ID NO:117); WVQQKPGQPPKRLIY (SEQ ID NO:118); WYQQKPGQPPKLLIS (SEQ ID NO:119); WYQQKPGQPPKLLIY (SEQ ID NQ:120);WYQQKPGQPPKRLIY (SEQ ID NO:121 ); WYQQKPGQPPKVLIY (SEQ ID NO:122); EASKLAS (SEQ ID NO:123); EASKVAS (SEQ ID NO:124); KASTLAS (SEQ ID NO:125); KTSTLAS (SEQ ID NO:126); RASTLAS (SEQ ID NO:127); RTSDLAS (SEQ ID NO:128); SASTLAS (SEQ ID NO:129); YASTLAS (SEQ ID NQ:130); GVPSRFKGSGFGTQFTLTISDVQCDDVATYYC (SEQ ID NO:131);GVPSRFKGSGSGTEFTLTVSDLECADAATYYC (SEQ ID NO: 132);GVPSRFSGSGSGTQFTLTISDLECDDAATYYC (SEQ ID NO:133);GVSSRFKGSGSGTDFTLTIRDLECADAATYYC (SEQ ID NO: 134);GVSSRFKGSGSGTQFTLTISDVQCDDAATYHC (SEQ ID NO:135);GVSSRFKGSGSGTQFTLTISDVQCDDAATYYC (SEQ ID NO: 136);GVSSRFKGSGSGTQFTLTISGVECADAATYYC (SEQ ID NO:137);GVSSRFKGSGSGTQLTLTISDLECADAATYYC (SEQ ID NO:138); AGGYSGSSDLCV (SEQ ID NO:139); AGGYSSVSDTA (SEQ ID NO:140); AGGYSSVTDTA (SEQ ID NO:141); AGGYSSVVDTA (SEQ ID NO:142); LGGYDSSSGDRWK (SEQ ID NO:143); LGIYEAGSDTS (SEQ ID NO:144);QCTYGSATTRIYGEA (SEQ ID NO:145); QCTYVDSSYIGG (SEQ ID NO:146); QQGYTGNNIDNP (SEQ ID NO:147); QSYDYGGGSYGNS (SEQ ID NO:148); FGGGTEVAVK (SEQ ID NO:149); FGGGTEVVVK (SEQ ID NQ:150).PD108 was digested with several different proteases under reducing and alkylated conditions. 140 pg of PD108 in 769 pL of volume was first reduced with DTT at a final concentration of 30 mM by heating at 95°C for 15 minutes. Then, the sample was divided into two fractions: 30% for cysteine modification with 2-Bromoethylamine hydrobromide (BEA) and 70% for iodoacetamide (IAA) alkylation. lAAwas added to a final concentration of 50 mM for 30 minutes in the dark at room temperature. The sample was precipitated by adding three times the sample volume of acetone and then incubated at -20°C for 1 hour. The sample was pelleted in a centrifuge at 23,000 x g for 10 minutes at 4°C; acetone was carefully decanted to avoid disrupting the pellet; then, the pellet was dried in a SpeedVac™ under low pressure. The pellet was resuspended in 10 pL of 4M urea at 37°C for 10 minutes to ensure complete dissolution, then diluted up to 100 pL with HPLC-grade water and separated into 20 pL across five tubes for the five digestions with LysC, trypsin, pepsin, chymotrypsin and AspN. Pepsin digestion was performed at 37°C for 15 minutes with shaking; the enzyme was then inactivated at 95°C for 3 minutes and dried in a SpeedVac™ under low pressure. The remaining sample digestions were performed overnight in a 37°C incubator, and the mixture was dried down using a SpeedVac™ under low pressure the following day. Finally, the reconstituted digest was analyzed by LC-MS / MS in a solution of 40 pL of 0.1% FA.BEA was added to the fraction set aside for cysteine modification at 320 pL of BEA in 100 mM tris pH 8, with 60 pL of 1 M tris pH added immediately and after each hour for 3 hours to maintain a reaction around neutral pH. The reaction was incubated at 25°C for 4 hours, then TCA was added to a final concentration of 20% and left at 4°C overnight. Proteins were precipitated by centrifuging at 17,000 rpm for 30 minutes at 4°C TCA was decanted, and the pellet was washed twice by adding 500 pL acetone and centrifuging at 17,000 rpm for 10 minutes; then, acetone was decanted from the pellet. The pellet was dried in a SpeedVac™ under low pressure and then reconstituted in 10 pL of 4M urea with shaking at 37°C for 10 minutes. The sample was diluted up to 40 pL with HPLC-grade water and then split evenly across two tubes. Digests were performed by adding 30 pL ammonium bicarbonate, pH 8 + 1 pg enzyme (trypsin for one, lysC for the other) and digested overnight in a 37°C incubator; then, dried down in a SpeedVac™ under low pressure the following day. Once acidified, 5 pg digested samples were loaded onto Evotips according to the manufacturer's instructions and run using Thermo Orbitrap™ Exploris 240 using a 44-minute gradient. All enzymes were obtained from Promega.Analysis of the sample yielded several regions de novo using Novor. These regions were determined based on the combination of a de novo approach and a database match of the protease digestion of the rabbit germline. The dominating heavy chain germlines were identified as IGHV1S69, IGHV1S40, IGHV1S44, and IGHV1S55, while the light chain germlines were IGKV1S50, 15, 17, 10, 32, and 1. No lambda chains were detected in the sample. A list of the few primary forms is shown in Table 8, with some mutations not found in the IMGT database. The light chain population was found to be more diverse than the heavy chain population.The initial search for CDR3 Heavy and Light for PD108 r1Table 8 displays the different areas for the light and heavy chains. This equates to a total of 21 ,600 potential combinations for the heavy chains and 573,440 possibilities for the light chains.To decrease the size of the database, a combinatorial database was generated for each individual and more exclusive CDR3 regions. In this example, CDR3H[3] = DGNSAYNSGVNL (SEQ ID NO:93) and CDR3L

[0010] = QSYDYGGGSYGNS (SEQ ID NO:148) were used, resulting in 3,600 heavy sequences (2*4*6*5*5*1*3) and 57,344 light sequence (8*8*7*8*8*1*2), respectively.Regarding the light chain, the number of combinations was manually reduced by testing the two different FR4s for the targeted sequence QSYDYGGGSYGNS (SEQ ID NO:148): Seq-a-b-c- d-e-10-1 : QSYDYGGGSYGNS (SEQ ID NO:148) + FGGGTEVAVK (SEQ ID NO:149) no overlap peptide was detected in the initial digest PD108A. However, for the sequence, Seq-a-b-c-d-e-10- 2: QSYDYGGGSYGNS (SEQ ID NO:148) + FGGGTEVWK (SEQ ID NQ:150), the peptide sequence QSYDYGGGSYGNSFGGGTEVWK (SEQ ID NO:151), fit several experimentalsequences found in the MS analysis. At 1164.5244amu as a 2+ is found with an intensity of 1 ,77e10. Thus, for the combinatorial database for the light chain, we have now:Seq-a-b-c-d-e-10-2Where "a," FR1 is any of the eight possibilities,"b," CDR1 any of the eight possibilities,"c" FR2 any of the seven possibilities,"d" CDR2 any of the eight possibilities,"e", FR3, any of the eight possibilities"f," CDR3

[0010] is fixed at a single possibility of QSYDYGGGSYGNS (SEQ ID NO: 148)"g," FR4[2] is fixed at a single possibility of FGGGTEVWK (SEQ ID NO: 150)There are 28,672 possibilities for the kappa light chain (8*8*7*8*8*1*2) and 3,600 possibilities for the heavy chain. Those two FASTA databases were then generated: for light chain Seq-a-b-c-d-e-10-2; for heavy chain: Seq-a-b-c-d-e-3-g.Two FASTA databases were used in conjunction with initial experimental sets of the experiment (7 different digestion conditions) which were searched using Novor cloud. The parameters for this search included 5 variable modifications of Cethyl(C), Pyro-Glu(Q), Carbamidomethyl(C), Deaminated(NQ), pyro-Glu(E), with the modification of Cethyl(C) corresponding to a modification similar to ethanolamine at 43.042199 amu on Cysteine. Mass tolerance was 5ppm for the precursors and 100ppm for the fragment, with 1% FDR. After searching the database, the protein list obtained from Novor cloud was not significantly reduced. For example, for light chain Seq-a-b-c-d-e-10-2, 26,026 proteins were reported from 28,672 sequences (less than 10% reduction of the database). For the heavy chain: Seq-a-b-c-d-e-3-g, 1212 sequences were reported from 3,600 sequences (a 2 / 3 reduction). These reported proteins represent about 33% of the entire heavy database and 91% of the entire light chain database. To reduce the list of potential candidates, all identified peptides from both searches were extracted and peptide redundancy was removed while keeping any instance of a similar peptide with different modifications. Subsequently, the list of candidate protein sequences was reduced by discarding sequences which showed a lower overlap between amino acid positions 6 to 120 for the heavy chains and from positions 6 to 110 for the light chain.Two candidate proteins rank 1stand 74thby Novo-Cloud based on peptides counts and show the highest level of “minimum score overlap”, MO score with a value of 14 and 6 respectively (shown in FIG. 12A). Their respective sequences read as follows:G001 with a minimum overlap score of 14: seq-2-4-6-1 -4-3-2G074 with a minimum overlap score of 6: seq-2-4-6-1 -4-3-3Both sequences show a high level of similitude with differences only within the “FR4” region.A FASTA database comprising of 68 heavy-chain protein candidates that had a minimum overlap score greater than 0 was generated. This database included the two potential candidate sequences differing only in their FR4 or J region, as well as an additional 66 sequences, all having a minimum overlap score of 3. These 68 proteins were selected from the 1 ,212 sequences reported by Novor cloud.The distribution of minimum overlap (MO) scores for the light chain, as shown in FIG. 12B, indicates that 263 proteins have the same MO score, an MO score of 10, which is also the highest score. Therefore, only this dataset was taken into consideration for the light chain to generate a FASTA file. This yielded 331 sequences used with the search engine pLink to increase confidence in pairing distant regions for both heavy and light chains together (68 candidates for the heavy chains and 263 candidates for the light chains). This approach aims to reduce the number of possible candidates by performing some additional orthogonal experiments. In that case, digestion under non-reduced conditions was performed to identify long-distance pairing within the selected candidates. This resulted in a smaller dataset which generated a FASTA file for potential heavy and light chain candidates to be searched using non-reduction digestion for distant region pairing in combination with pLink. This method was employed to increase confidence in pairing distant regions (based on disulfide peptide association) and to reduce the possible number of candidate sequences even further.For non-reduction digestion analysis with pLink, the sample preparation consisted of two conditions. Condition 1 involved 20 pg of the polyclonal antibody PD108 diluted to 20 pL in 10 mM phosphate buffered saline (PBS) with two mM N-ethylmaleimide, then incubated at 37°C for 2 hours with shaking. The sample was then dried under low pressure in a SpeedVac™, resuspended in 25 pL 8M urea in 100 mM tris buffer, adjusted to pH 6.5, and digested overnight with LysC at a ratio of 1 :20 (protease to protein ratio) at 37°C in an incubator. The following day, the sample was diluted with 100 mM tris buffer, pH 6.5, to a final urea concentration of 2M and digested with AspN at a ratio of 1 :20 (protease to protein ratio) for 4 hours at 37°C in an incubator. Condition 2 involved 20 pg of PD108, dried completely under low pressure, then reconstituted in 20 pL 8M urea in 100 mM tris buffer, adjusted to pH 6.5, to which NEM was added to a final concentration of 2 mM and incubated at 37°C for 2 hours with shaking. LysC was added at a ratio of 1 :20 (protease to protein ratio) at 37°C in an incubator for overnight digestion, followed by 4- hour digestion with AspN at a ratio of 1 :20 (protease to protein ratio) at 37°C in an incubator. After protease digestion, samples in both conditions were dried completely under low pressure and reconstituted in 40 pL 0.1% formic acid. Then, 2.5 pL of the digested samples were loaded onto Evotips according to the manufacturer's instructions and run using a Thermo Orbitrap Exploris 240 on a 44-minute gradient. The identification of the S-S bridge containing peptides was performed using pLink v2. The software identification parameters were set as follows: Disulfidebond (HCD-SS); Enzyme: LysC-AspN, Try, orTry-GluC; Peptide mass: 300-9000; Peptide length: 3-90; Fixed modification: Gln->pyro-Glu; and Variable modifications: "Nethylmaleimide[C]," "Oxidation[M]," "Deamidated[N]," "Deamidated[Q]," and "Acetyl[ProteinN-term]." The protein database contained the sequences of interest for the pAb mix of interest and the corresponding reversed sequences. Table 9 illustrates some of the identified peptides from the non-reduced experiment. As the peptides can correspond to several possible protein sequences, the peptide assignment was restricted to a specific region (FR1 , CDR1 , FR2, etc.), with the bracket illustrating one of the two peptides linked by S-S. In some cases, the information obtained from the peptide was insufficient to identify a given region; thus, a range was proposed as shown in table 6 from the previous example. Peptides from the FR4 region were not reported as they were not identified.For peptides #1 and #2 of the light chain, CDR1 [4] could be paired with CDR3

[0010] , thus limiting some of the FR2 and FR3 possibilities.Table 9CQASQSISSYLAWYQQK (SEQ ID NO:152); DAATYYCQSY(SEQ ID NO:153);CQASQSISSYLAWYQQKPGQPPK (SEQ ID NO:154);QSVEESGGRLVTPGGSLTLTCTVSGFSLNNYPMAWVRQAPGK (SEQ ID NO:155); DTATYFCAR (SEQID NO:156); QSVEESGGRLVTPGGSLTLTCTVSGFSLNTYPMGWVRQAPGK (SEQ ID NO:157);QSVEESGGRLVTPGGSLTLTCTVSGFSLSSYGVSWVRQAPGK (SEQ ID NO: 158); DTATYFCTR (SEQID NO:159); QSVEESGGRLVTPGGSLTLTCTVSGFSLSTYSMSWVRQAPGK (SEQ ID NO:160);QSVEESGGRLVTPGTPLTLTCTVSGFSLNNYPMAWVRQAPGK (SEQ ID NO: 161 );QSVEESGGRLVTPGTPLTLTCTVSGFSLNNYPMGWVRQAPGK (SEQ ID NO:162);QSVEESGGRLVTPGTPLTLTCTVSGFSLNNYPMGWVRQAPGK (SEQ ID NO:163);QSVEESGGRLVTPGTPLTLTCTVSGFSLNTYPMGWVRQAPGK (SEQ ID NO: 164);QSVEESGGRLVTPGTPLTLTCTVSGFSLSSYAMGWVRQAPGK (SEQ ID NO: 165);QSVEESGGRLVTPGTPLTLTCTVSGFSLSTYSMSWVRQAPGK (SEQ ID NO:166); DTATYFCAT (SEQ ID NO: 167); QSVEESGGRLVTPGGSLTLTCTVSGFSLSTYAMGWVRQAPEK (SEQ ID NO: 168); CQASQSVYNNDRLSWVQQK (SEQ ID NO:169).The light chain has a potential candidate sequence: Seq-a-4-(4-7)-d-(2,3,4,6,7,8)-10-2, which can be interpreted as having any of eight possible FR1 , CDR1 [4], FR2 with any of 4, 5,6, 7, any of eight possible CDR2, FR3 with any of 2, 3, 4, 6, 7, 8, CDR3

[0010] and FR4[2] (as only evidence of peptides for CDR3

[0010] and FR4[2] has been found). This gives a FASTA dataset of 8x1x4x8x6x1x1 = 1536 sequences. For the heavy chain, no CDR3 peptides were identified; however, one of the most promising candidates is Seq-2-4-6-d-4-f-g, which gives a FASTA dataset of 5x6x3 = 90 sequences to help confirm this heavy chain. (See Table 8 for details).A repeat search was performed on the Novor cloud platform using the following conditions: Condition 1 : A smaller FASTA combinatorial dataset was refined from the NR dataset experiment. The Light chain sequence was Seq-a-4-(4-7)-d-(2,3,4,6,7,8)-10-2 (1536 candidates) and the Heavy chain sequence was Seq-2-4-6-d-4-f-g (90 candidates).Condition 2: A larger experimental LC-MS dataset was used, which included the initial dataset plus some additional LC-MS analyses acquired in "Middle-Down" mode (including the initial different digestion conditions plus a few selected additional samples such as AspN, LyC,trypsin, and LyC with Cysteine modified with bromoethylamine, chymotrypsin, and pepsin run in the middle-down proteomics mode to capture longer peptides).The parameters on Novor-Cloud were the same as those used in the first round of analysis. The search was conducted in two steps: first, the combinatorial database from Condition 1 was used and the data acquired under standard Evosep conditions was searched (standard proteomics in bottom-up mode). Then, the search was repeated on the middle-down dataset. The lists of peptides found from both searches was extracted, and only the non-redundant ones were kept, as previously described.Regarding the heavy chain, Novor cloud was used to generate a list of 17 sequences from the 90 FASTA sequences provided. The minimum overlap score was evaluated from amino acid positions 6 to 119. Table 10 ranks these 17 proteins identified by their minimum overlap score. All of the 17 selected proteins have a “d” value of “1” (in seq-a-b-c-d-e-f-g thus we have seq-2-4- 6-1-4-f-g). Although a minimum score overlap value helps eliminate protein sequences with no overlap, it does not provide a detailed overview of the overall overlap score distribution. FIG. 13 displays the three best hits sequences from Table 10 starting with seq-2-4-6-1-4; the sequence seq-2-4-6-1 -4-3-2 has a higher quality overlap from the CDR3 and FR4 included and was, therefore, kept as a potential candidate for the heavy chain.Table 10Regarding the light chain, from the 1 ,536 fasta sequences searched, Novor cloud kept 1 ,320 sequences. Of these, 1099 had no zero gaps. The top 3 candidates had the same MO score of 16, and are shown overlaid in FIG. 14:G0740 (Seq- 1-4-5-5-8-10-2),G1080 (Seq-1 -4-6-3-8-10-2) andG1143 (Seq- 1-4-5- 1-8- 10-2).All 3 sequences have the same minimum Overlap score scores at aa position 89 (FIG. 14). However, plotting the overall overlap score revealed that Seq-1-4-5-5-8-10-2 (G0740) had a higher overlap score for regions covering the CDR1 , FR2, CDR2, and FR3 (illustrated by the arrow in FIG. 14). This sequence was therefore chosen as the best candidate and paired with a heavy chain using experimental and bioinformatical-based strategies as described in PCT / CA2022 / 051194.Recombinant antibodies were produced, and their affinities were tested against the natural polyclonal antibody. FIG. 15 shows the ELISA curve for the recombinant form R1 vs. the natural form PD108 and a negative control. The recombinant was found to have a slightly higher affinity than the original natural pAbs, likely due to the difference in purity between the recombinant and the natural polyclonal mixture.Example 4: Usage of crosslink to help long-distance sequence stretch within a single chain from a mixtureIn this example, the usage of crosslinkto help long-distance sequence stretch within a single chain from a mixture (an artificial mixture of rabbit monoclonal antibodies (internal standard naming P17 and P18), was investigated. The antibody mixture was prepared in equal amounts (40 pg, 1 pg / pL each). BS3 crosslinker from ThermoFisher (cat # A39266) was prepared to 50 mM in 25 mM Sodium Phosphate, pH 7.4 and added to the antibody at 125 and 150 molar excess. The reaction was allowed to proceed at 25°C for 1 hour, then neutralization with Tris pH 8 (final Tris concentration of 60 mM) at room temperature for 30 minutes. 20 pg of each condition was then diluted to 50 pL using HPLC grade water and reduced using DTT (final concentration of 30 mM) and heating at 95°C for 15 minutes. Subsequently, IAA was added to a final concentration of 50 mM for alkylation at room temperature for 30 minutes in the dark. The sample was then precipitated by adding three times the sample volume of acetone at -20°C for 1 hour, followed by centrifugation at 23,000 x g for 10 minutes at 4°C. The pellet was dried using a SpeedVac™ under low pressure and then resuspended using 4 pL 4M urea at 37°C with shaking for 10 minutes. After digestion with 1 pg pepsin, 2 pL 1 N HCI to acidify, and diluting up to 50 pL with HPLC grade water, pepsin was inactivated by heating the sample at 95°C for 3 minutes. The sample was dried down in a SpeedVac™ under low pressure and then reconstituted in 40 pL 0.1% FA. Five pg wasloaded on Evotips and run on Thermo Orbitrap™ Exploris 240 using an 88-minute gradient method for data analysis with pLink.As shown in T able 11 , the highest number of sequences was observed at BS3 molar excess of 150x. In the mixture of two relatively similar antibodies, specific distant regions were paired on both the heavy and light chains, including CDR1 with CDR2 (peptides 1 , 3, 14 and 18), CDR1 with CDR3 (peptide #8 and #9). Furthermore, different framework regions were as well paired with other frameworks or CDRs regions, with up to five different regions assembled from two linked peptides (e.g., peptide #3 which contains a single peptide from FR1 , CDR1 , FR2, connected to another peptide from CDR2, and FR3). Interchain pairing was also observed(peptides 20 and 21). No crosslink between different antibodies was observed.Table 11Notes1) BS3 ME: Molar ExcessASGVPSRFKGSGSGTE (SEQ ID NO:170); IKCQASQSIY (SEQ ID NO:171); IYDASKL (SEQ ID NO:172); ETGVPSRFKGSGSGTRF (SEQ ID NO:173); ISCQSSQSVNKNDLSW (SEQ ID NO:174); ETGVPSRFKGSGSGTRF (SEQ ID NO: 175); IYEASKL (SEQ ID NO: 176); IKCQASQSIY (SEQ ID NO:177); KGSGSGTE (SEQ ID NO:178); KGSGSGTEFT (SEQ ID NO:179); TCKASGFD (SEQ ID NO:180); CGKDLGL (SEQ ID NO:181 ); TCKASGFDF (SEQ ID NO:182); FCGKD (SEQ ID NO:183); ISCQSSQSVNKNDL (SEQ ID NO:184); KGSGSGTRF (SEQ ID NO:185); ISCQSSQSVNKNDLS (SEQ ID NO:186); ISCQSSQSVNKNDLSW (SEQ ID NO:187); KGSGSGTRFTL (SEQ ID NO:188); YEASKL (SEQ ID NO:189); IYEASKL (SEQ ID NO:190); YQQKPGQPPKLL (SEQ ID NO:191 ); LIYDASKL (SEQ ID NO:192); YFCGKDLGL (SEQ ID NO:193); LIYEASKL (SEQ ID NO:194); KGSGSGTRFTLT (SEQ ID NO:195).This experiment shows that the usage of crosslink followed by protease digestion applied to a mixture of similar antibodies permits the pairing of distant regions within individual chains within that mixture of similar antibodies.Example 5: Sequencing of a human IgG anti-Receptor Binding Domain isolated from plasma from an individual vaccinated against COVID-19For this example, sequence-specific antibodies binding to the antigen produced by vaccination in humans were enriched to assess the capacity of the receptor binding domain (RBD)-containing COVID-19 vaccine to generate antibodies against the RBD of the SARS-CoV2 virus in the human host. These antigen-specific antibodies were purified and sequenced using a de novo proteomics-based approach. Following sequencing, recombinant forms were generated, and ELISA testing was performed on the recombinant forms. The main element of this particular example is to show that a complex mix of intact Fab2antibody protein fragments specific to the RBD antigen can be separated using a native gel approach and allow distant region assembly based on the fraction profile. From the mixture of antigen-specific antibodies, the proposed method is a 2-step approach comprising: 1) digesting the polyclonal with several proteases (i.e., a bottom-up approach which allows generating contig) and 2) separating the intact antibody or Fab2, to assemble the different contig / distant region.Three samples were collected from 3 different healthy Donors. Data from a single female donor (named “522”) are presented. Blood was taken about 2 months after the subject received a 3rd shot of the Moderna Spikevax® COVID-19 vaccine.IgG enriched against the antigen:Total IgG was enriched from 3 mL human serum from the subject vaccinated for COVID. Briefly, 3 mL of settled protein G agarose resin (Genscript Cat# L00209) in a 20 mL gravity flow Biorad column (cat# 7321010EDU) was equilibrated by washing twice with 15 mL 10 mM phosphate buffered saline (PBS). Serum was prepared for enrichment by centrifuging at 23,000 RCF for 10 min at 4°C to precipitate debris. The 3 mL serum was combined with 9 mL 10 mM PBS, then filtered. Filtered serum was passed over protein G agarose resin using gravity flow, then flow through was passed over the same Protein G column twice to bind IgG. Protein G resin was washed three times using 10 mL 10 mM PBS then eluted using 12.5 mL glycine buffer 0.1M pH 2.5. Eluted IgG was then concentrated, and buffer was exchanged into 10 mM PBS using 30 kDa Amicon™ filter (Sigma cat#UFC803024) and the total amount of IgG captured was found to be 22.2 mg.Antigen enrichment was performed to enrich anti-RBD antibodies by coupling 0.4 mg SARS-CoV-2 spike protein receptor binding domain (RBD) to 54 pL streptavidin coated agarose beads (Sigma cat#GE17-5113-01) for 1 hour at 4°C with tumbling. Biotinylation of RBD was first performed using a 20-times excess of 20 mM biotin (Sigma cat#A39259) reconstituted in water,followed by a buffer exchange into 10 mM PBS using a 3 kDa Amicon™ filter (Sigma cat#UFC500396) to remove excess biotin. Following coupling of biotinylated RBD to streptavidin resin, three washes were performed using 0.4 mL 10 mM PBS to remove any uncoupled RBD then 20 mg of the total IgG purified above was added and incubated for 1 hour at 4°C with tumbling. Non-specific binders were removed by washing twice with 0.4 mL 10 mM PBS followed by a wash with 0.4 mL CHAPS 0.5% in 10 mM PBS and then 6 washes with 0.4 mL 10 mM PBS. Anti-RBD antibodies were eluted twice using 0.4 mL 0.1 M glycine buffer pH 2.5 incubated for 5 min at room temperature and combined, then neutralized using 0.2 mL 1 M Tris pH 8 resulting in 111 pg anti-RBD antibodies, which were given an internal sample name of PD124.In-solution Digestion: An in-solution digestion was performed on 25 pg of the sample PD124 by concentrating the sample to 100 pL using centrifugation under low pressure (i.e., Speedvac™), followed by reduction using dithiothreitol (DTT) at a final concentration of 30 mM at 95°C for 15 minutes. The sample was then split into two, 5 / 7th was alkylated with lodoacetamide (IAA) at a final concentration of 50 mM at room temperature in dark conditions for 30 minutes, and 2 / 7th was treated with 2-bromoethylamine hydrobromide (BEA) at a final concentration of 0.5 M for 4 hours at 25°C with shaking, adding 10 pL of 1 M Tris pH 8 hourly (40 pL total) to maintain neutral pH during reaction. The sample treated with IAA was then precipitated by adding three times the sample volume of acetone at -20°C for 1 hour, followed by centrifugation at 23,000 x g for 10 minutes at 4°C. Acetone was decanted and the pellet was dried under low pressure in Speedvac™, followed by reconstitution with 10 pL 4M urea which was incubated at 37°C for 10 min with shaking to reconstitute the pellet. The sample was then split across 5 tubes, 30 pL of 50 mM ammonium bicarbonate was added to 4 of these for digestion with Trypsin, LysC, AspN and Chymotrypsin at a 1 :20 ratio of protease:protein. These samples were left to digest overnight in a 37°C incubator. For the 5th digestion, 30 pL of HPLC grade water was added as well as 2 pL of 1 N HCI and pepsin was added at a 1 :20 ratio of protease:protein. Pepsin digest occurred at 37°C for 15 minutes with shaking, followed by a 3 min pepsin inactivation at 95°C. Digests were then dried under low pressure in the Speedvac™.BEA-treated samples were precipitated by adding trichloroacetic acid to a final volume of 20% then incubating overnight at 4°C. The following day, samples were pelleted with centrifugation at 17,000 RPM for 30 minutes at 4°C then pellets were washed twice with 500 pL 80% acetone with centrifugation at 23,000 RPM for 10 min for each wash step. Pellets were dried under low pressure using Speedvac™ and then reconstituted in 4 pL 4 M urea at 37°C for 10 min with shaking then split across 2 tubes and digested in ammonium bicarbonate at a final concentration of 30 mM with Trypsin and LysC at a ratio of 1 :20 protease: protein. Digestions were incubated at 37°C overnight. All proteases used for PD124 digests were obtained from Promega.Digested samples were dried to completion in Speedvac™ under low pressure, then resuspended in 40 pL 0.1 % formic acid. 2.5 ug was loaded on Evotips according to manufacturer’s instructions for each digestion and run on Orbitrap™ 240 Exploris using the 30 samples per day method (44min method) on a 15cm PepSep column. Data were acquired at 60,000 resolution for 400-2000 m / z precursors with standard AGO target and maximum injection time set to auto. An intensity threshold of 2.5e4 was applied, and charge states 2-8 were included. Dynamic exclusion was set to 15s. MS / MS fragmentation was induced with fixed 30% HOD, detected by the Orbitrap™ with 7500 resolution with a maximum injection time of 50ms.Native gel. PD124 was separated on gel using BioRad precast 7.5% polyacrylamide gels (cat#4561024) with and without IdeS treatment. For IdeS treatment, 50 ug of PD124 was incubated at 37°C with 50 units of IdeS from Promega for 1 hour then dried in a Speedvac™ under low pressure to reduce volume to 30 pL. Nativepage 4x buffer (Thermo cat#BN20032) was added in a 4:1 ratio. For the undigested sample, only 25 pg was loaded on gel following the addition of NativePage 4x buffer in a 4:1 ratio.Running buffer was made by diluting 10X stock Tris / Glycine Buffer (BioRad cat#1610771) to 1X using Milli-Q water. The gel was run using a BioRad power bank set to 130V for 180 min, then stained with Coomassie brilliant blue (BioRad, cat#1610436) for 30 min and destained with BioRad destain (cat#1610438) overnight.The gel was cut into 12 bands (FIG. 16). Bands / fractions 4, 7, 9, and 10 have a higher protein abundance and were further split into 2 or 3 pieces. Trypsin, pepsin, and chymotrypsin were used to digest the first, second and third (when available) of each band using a standard ingel digestion protocol. Bands were dehydrated with 200 pL 100 mM tetraethylammonium bicarbonate (TEAB) and AON in a 1 :1 ratio, which was allowed to sit for 5 minutes before aspirating to remove and then further dehydrating bands with 200 pL AON for 30 seconds. After removing AON, bands were left at room temperature to air dry, then reconstituted in 25 mM DTT in 100 mM TEAB to reduce at 56°C for 30 min with shaking. Following this, bands were cooled to room temperature, DTT was removed and then 55 mM IAA was added to alkylate for 30 min at room temperature in dark. IAA was removed, bands were washed twice with 0.4 mL HPLC grade water then dehydrated with 200 pL 100 mM TEAB and AON in a 1 :1 ratio, for 5 minutes before aspirating to remove and then further dehydrating bands with 200 pL AON for 30 seconds. T rypsin and Chymotrypsin were diluted to 6 ng / pL in 100 mM TEAB, then 100 pL was added to gel pieces for digestion. Pepsin was diluted to 20 ng / pL and acidified using 1 N HCI, then 100 pL was added to gel pieces for digestion. All digestions were performed at 37°C overnight. The following day, supernatant was collected in new tubes, gel pieces were dehydrated to extract additional peptides using 100 pL 60% ACN in 40% 0.1% FA, then sonicated for 30 minutes. The supernatant from this extraction was combined with the digestion supernatant and dried down under low pressure in Speedvac™.Following gel separation, bands were excised using a razor for both IdeS digested and undigested lanes and washed twice with 200 pL HPLC grade water. Intense bands were cut into multiple pieces to perform multiple enzyme digests.Dried samples were resuspended in 40 pL 0.1% FA and then 100% of sample was loaded on Evotips according to manufacturer’s instructions and run on an Orbitrap™ 240 Exploris using 30 samples per day (44min method) on a 15cm PepSep column. Data was acquired at 60,000 resolution for 400-2000 m / z precursors with standard AGO target and maximum injection time set to auto. An intensity threshold of 2.5e4 was applied, and charge states 2-8 were included. Dynamic exclusion was set to 15s. MS / MS fragmentation was induced with fixed 30% HOD, detected by the Orbitrap™ with 7500 resolution with a maximum injection time of 50ms. The IdeS digested sample in FIG. 16 shows clear bands, which suggests a fractionation of the polyclonal mixture of antibodies.Data analysis. The MS / MS spectra in the bottom-up data were de novo sequenced with Novor software. Candidate contig sequences covering each CDR area were assembled by combining overlapping de novo peptides. The assembly can be continued whenever the ambiguity for the two peptides’ overlap is low, for example, when the overlap involves multiple (>1) amino acids that are mutations from germline sequences. Table 12 shows some example contig sequences obtained from the assembly.Table 12The contig sequences covering distant CDRs were assembled with the assistance of the quantification data obtained from the native gel-based separation experiment. For each unique peptide in a contig, the quantity (peak area) of the peptide in each fraction is calculated with MaxQuant software by using the label-free quantification method. A normalized quantification vector for a peptide is obtained by dividing its quantity in each fraction by the total in all fractions. The normalized quantification vectors of the peptides for each contig in Table 12 are plotted as curves in FIGs. 17A-C. It is already clear from FIGs. 17A-C that HCDR1-C02, HCDR2-C01 andHCDR3-cO1 share very similar intensity curves with each other. This strongly suggests the three contig sequences belong to the same antibody protein and should be paired together. The same conclusion can be also made by calculating a similarity score as follows. The similarity score between a pair of contigs is the average Pearson correlation coefficient between every pair of unique peptides from the two contigs, respectively. Among the HCDR1 and HCDR2 contigs, HCDR1-C02 and HCDR2-C01 have the highest similarity scores to HCDR3-C01 , respectively. This also can conclude that these three contig sequences should be paired together. As in Example 1 , using separation of intact antibody combined with a digest of the entire polyclonal mixture allow for assembling distant regions.Finally, the three selected contigs HCDR1-C02, HCDR2-C01 , and HCDR3-C01 were aligned to the variable germline genes and the gaps between them were filled by adding other de novo peptides that connect them. FIG. 18 illustrates the process. The proper assignment of the isobaric Leucine and Isoleucine were determined by their w-ions in separate EThcD experiments using C- terminal chemistry as described in PCT / CA2019 / 051870). After adding the constant region, the following final sequence was obtained: >PD124-R5-heavy (SEQ ID NO:203) EVQLVESGGDLVQPGGSLRLSCAASGFTFSNYDMHWVRQVTGKGLEWVSGIGKDGDTYYLGSVKGRFAIS RDNAKNSLYLQMNSLRAGDTALYYCARVGTTGYDLYGMDVWGQGTTVTVSSTSTKGPSVFPLAPSSKSTS GGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSWTVPSSSLGTQTYICNVNHKP SNTKVDKKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVWDVSHEDPEVKFN WYVDGVEVHNAKTKPREEQYNSTYRWSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREP QVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSR WQQGNVFSCSVMHEALHNHYTQKSLSLSPGK (SEQ ID NO : 203 )A similar procedure also obtained a light chain sequence: >PD124-R5-light (SEQ ID NO:204) SYELTQPPSVSVSPGQTARITCSGNVFPRQYAYWYQQKPGQAPVLLIYKDSERPSGIPERFSGSGSGTTV TLTITGVQAEDEADYYCQSGDSGGWVFGGGTKLTVLGQPKAAPSVTLFPPSSEELQANKATLVCLISDFY PGAVTVAWKADSSPVKAGVETTTPSKQSNNKYAASSYLSLTPEQWKSHRSYSCQVTHEGSTVEKTVAPTE CSThe pair of heavy and light chain sequences were paired up by comparing their quantification changes in the separation experiments in a similar way as the contigs from different CDRs were paired (the heavy and light chain pairing is based on a procedure using experimental and bioinformatical-based strategies as described in PCT / CA2022 / 051194). The antibody was expressed recombinantly in HEK293 cells and demonstrated similar binding affinity to the human anti-RBD pAb extracted from the serum.Although the present invention has been described hereinabove by way of specific embodiments thereof, it can be modified, without departing from the spirit and nature of thesubject invention as defined in the appended claims. In the claims, the word "comprising" is used as an open-ended term, substantially equivalent to the phrase "including, but not limited to". The singular forms "a", "an" and "the" include corresponding plural references unless the context clearly dictates otherwise.REFERENCES1. https: / / www.marketdataforecast.com / market-reports / antibodies-market2. Wang et al.,_Back to the future: recombinant polyclonal antibody therapeutics. Curr Opin Chem Eng. 2013 Nov;2(4):405-415. doi: 10.1016 / j.coche.2013.08.005.3. Cheung, W., Beausoleil, S., Zhang, X. et al. A proteomics approach for the identification and cloning of monoclonal antibodies from serum. Nat Biotechnol 30, 447-452 (2012). https: / / doi.Org / 10.1038 / nbt.21674. Wine et al., Molecular deconvolution of the monoclonal antibodies that comprise the polyclonal serum response, PNAS 110 (8) 2993-2998 (2013).5. Gilchuk et al., Proteo-Genomic Analysis Identifies Two Major Sites of Vulnerability on Ebolavirus Glycoprotein for Neutralizing Antibodies in Convalescent Human Plasma. Front. Immunol., 15 July 2021 , Volume 12 - 2021 https: / / doi.org / 10.3389 / fimmu.2021.7067576. Nesvizhskii et al., Interpretation of shotgun proteomic data: the protein inference problem. Mol Cell Proteomics. 2005 Oct;4(10): 1419-40. doi: 10.1074 / mcp.R500012-MCP200. Epub 2005 Jul 11.7. Guthals et al. De Novo MS / MS Sequencing of Native Human Antibodies. J. Proteome Res. 2017, 16, 1 , 45-54. October 25, 2016, https: / / doi.org / 10.1021 / acs.jproteome.6b006088. Fan et al. Using pLink to Analyze Cross-Linked Peptides. Curr Protoc Bioinformatics. 2015 Mar 9:49:8.21.1-8.21.19. doi: 10.1002 / 0471250953.bi0821s49.9. Yu, F., Li, N., and Yu, W. (2016). ECL: an exhaustive search tool for the identification of cross-linked peptides using whole database. BMC Bioinformatics, 17(1), 110. Rinner et al., Identification of cross-linked peptides from large sequence databases. Nat Methods. 2008 Apr; 5(4): 315-318. Published online 2008 Mar 9. doi: 10.1038 / nmeth.1192.11 . Leitner A, Walzthoeni T, Aebersold R "Lysine-specific chemical cross-linking of protein complexes and identification of cross-linking sites using LC-MS / MS and the xQuest / xProphet software pipeline." Nat Protoc 2014; 9(1): 120-37.12. Chu F, Baker PR, Burlingame AL, Chalkley RJ. Finding chimeras: a bioinformatics strategy for identification of cross-linked peptides. Mol Cell Proteomics. 2010;9:25-31. doi: 10.1074 / mcp.M800555-MCP200.13. Trnka MJ, Baker PR, Robinson PJ, Burlingame A, Chalkley RJ. Matching cross-linked peptide spectra: only as good as the worse identification. Mol Cell Proteomics. 2014;13(2):420- 34. doi: 10.1074 / mcp.M 113.034009.14. Hoopmann MR, Zelter A, Johnson RS, Riffle M, MacCoss MJ, Davis TN, Moritz RL. Kojak: efficient analysis of chemically cross-linked protein complexes. J ProteomeRes. 2015;14(5):2190-198. doi: 10.1021 / pr501321 h.15. Netz et al. OpenPepXL: An Open-Source Tool for Sensitive Identification of Cross- Linked Peptides in XL-MS. Mol Cell Proteomics. 2020 Dec; 19(12):2157-2168. doi:10.1074 / mcp.TIR120.002186. Epub 2020 Oct 16.16. Pirklbauer et al. MS Annika: A New Cross-Linking Search Engine. J. Proteome Res.2021 , 20, 5, 2560-2569. Publication Date: April 14, 2021. https: / / doi.Org / 10.1021 / acs.jproteome.0c0100017. Chen et al., A high-speed search engine pLink 2 with systematic evaluation for proteome-scale identification of cross-linked peptides. Nature Communications, volume 10, Article number: 3404 (2019))18. Yang, B., Wu, YJ., Zhu, M. et al. Identification of cross-linked peptides from complex samples. Nat Methods 9, 904-906 (2012). https: / / doi.org / 10.1038 / nmeth.2099.19. PCT application No. PCT / CA2022 / 051194, published as WO 2023 / 010219.

Claims

WHAT IS CLAIMED IS:1 . A method for determining the amino acid sequence of one or more antibodies or antibody chains present in an antibody mixture, the method comprising:(A) generating a pool of antibody-derived peptide sequences by:(i) contacting a plurality of samples from the mixture with a reducing agent;(ii) optionally contacting the plurality of samples with an agent that either prevent disulfide bridge formation or modifies cysteine residues into lysine analogs;(iii) contacting the plurality of samples with one or more proteases and / or chemical proteolytic agents to obtain antibody-derived peptides, wherein each of the sample is contacted with a different protease, chemical proteolytic agent or combination thereof, thereby obtain a plurality of antibody-derived peptide digests;(iv) determining both the amino acid sequences and intensity of the short antibody- derived peptides present in the antibody-derived peptide digests by mass spectrometry using a de novo sequencing approach, wherein the short antibody- derived peptides have a length of less than 50 amino acids;(v) assigning the antibody-derived peptide sequences to a specific complementary determining region (CDR1 , CDR2 or CDR3) or framework region (FR1 , FR2. FR3 or FR4);(vi) generating a library of candidate antibody chain sequences in silica by combining the antibody-derived peptides sequences from different antibody regions identified in (v);(B) performing at least one of (a) to (e) to identify the amino acid sequence of one or more antibodies or antibody chains present in the antibody mixture from the library of candidate antibody chain sequences:(a)(i) optionally contacting the sample with an immunoglobulin (Ig) domain isolation protease to obtain Fab, F(ab’)2and Fc fragments; optionally removing the Fc fragments (e.g., using protein A / G beads);(ii) submitting the sample from the mixture to a separation step to separate the antibodies, antibody chains or antibody fragments (Fab, F(ab’)2and Fc) present in the mixture in a plurality of fractions;(iii) optionally contacting a sample from the antibody mixture with a reducing agent and / or with an agent that modifies cysteine residues either to prevent disulfide bridge formation of to modify them into lysine analogs;(iv) contacting the fractions with or more proteases and / or chemical proteolytic agents to obtain digested fractions comprising antibody-derived peptides, wherein the peptides correspond to different regions of the antibodies or antibody chains;(v) submitting the digested fractions to mass spectrometry (MS) to obtain MS and / or MS / MS spectra;(vi) analyzing the MS and / or MS / MS spectra with a proteomic search engine using the pool of antibody-derived peptide sequences obtained in (A); and(vii) assembling the antibody-derived peptides issued from the same antibody or antibody chain based on their co-elution profile across the different fractions;(b)(i) optionally incubating a sample from the antibody mixture with an agent that modifies free cysteine residues;(ii) incubating the sample under denaturating, non-reducing conditions;(iii) contacting the sample with one or more proteases and / or chemical proteolytic agents to obtain digested antibody-derived peptides, wherein the digested antibody- derived peptides comprise dimers of peptides from distant antibody regions linked by a disulfide bridge;(iv) submitting the digested fractions to mass spectrometry (MS) to obtain MS and / or MS / MS spectra;(v) analyzing the MS and / or MS / MS spectra with a proteomic search engine suitable for identifying cross-linked peptides using the antibody-derived peptide sequences identified in (A);(c)(i) incubating a sample from the antibody mixture with a protein cross-linking agent;(ii) incubating the sample under denaturating, reducing conditions;(iii) optionally contacting the sample with an agent that modifies cysteine residues either to prevent disulfide bridge formation or to convert cysteine into lysine analogs;(iv) contacting the sample with one or more proteases and / or chemical proteolytic agents to obtain digested antibody-derived peptides, wherein the digested antibody-derived peptides comprise cross-linked dimers of peptides from distant antibody regions;(v) submitting the digested fractions to mass spectrometry (MS) to obtain MS and / or MS / MS spectra;(vi) analyzing the MS and / or MS / MS spectra with a proteomic search engine suitable for identifying cross-linked peptides using the antibody-derived peptide sequences identified in (A);(d)(i) generating a library of candidate antibody chain sequences in silica by combining the antibody-derived peptides sequences from different antibody regions identified in (A)(v);(ii) determining, by mass spectrometry, both the amino acid sequences and intensity of the long antibody-derived peptides present in the plurality of antibody-derived peptide digests from (A)(iii) , wherein the long antibody-derived peptides have a length of more than 50 amino acids;(iii) comparing the long antibody-derived peptide sequences with the library of candidate antibody chain sequences to identify antibody chain sequences present in the antibody mixture;(e)(i) assessing the overlap between the antibody-derived peptides present in the sample, wherein a high level of overlap is indicative that the antibody-derived peptides belong to the same antibody or antibody chain.

2. The method of claim 1 , wherein the separation step comprises a chromatography or gel separation.

3. The method of claim 2, wherein the chromatography is Hydrophobic Interaction Chromatography (HIC), and wherein the gel separation is native gel, isoelectric focusing (IEF) gel, or 2D gel separation.

4. The method of any one of claims 1 to 3, wherein the separation step is performed under non-reducing conditions.

5. The method of any one of claims 1 to 4, wherein the protease is pepsin, trypsin, chymotrypsin, AspN, LysC, GluC or any combinations thereof.

6. The method of any one of claims 1 to 5, comprising incubating the sample from the antibody mixture with a protein cross-linking agent.

7. The method of claim 6, wherein the protein cross-linking agent comprises bis(sulfosuccinimidyl)suberate (BS3).

8. The method of any one of claims 1 to 7, wherein the protein sequence database search engine is Mascot, Sequest, Novor-Cloud or Maxquant.

9. The method of any one of claims 1 to 8, which comprises contacting the sample with an agent that modifies cysteine residues into lysine analogs.

10. The method of claim 9, wherein the agent that modifies cysteine residues to prevent disulfide bridge formation comprises iodoacetamide, and / or the agent that modifies cysteine residues into lysine analogs comprises 2-Bromoethylamine hydrobromide (BEA).11 . The method of any one of claims 1 to 10, which comprises contacting the sample with an agent that prevents cysteine residues from forming disulfide bonds.

12. The method of claim 11 , wherein the agent that prevents cysteine residues from forming disulfide bonds comprises N-ethylmaleimide.

13. The method of any one of claims 1 to 12, wherein the Ig domain isolation protease comprises IdeS and / or IdeZ.

14. The method of any one of claims 1 to 13, wherein the short antibody-derived peptides have a length of 5 to 20 amino acids.

15. The method of any one of claims 1 to 14, wherein the long antibody-derived peptides have a length of 25 to 100 amino acids.

16. The method of claim 15, wherein the long antibody-derived peptides have a length of 40 to 80 amino acids.

17. The method of any one of claims 1 to 16, wherein the level of overlap between two antibody- derived peptides is determined by calculating an overlap score, and wherein an overlap between amino acids located at the amino and carboxy-terminal ends of the antibody-derived peptides is given a lower score relative to an overlap between amino acids located at internal positions in the antibody-derived peptides.

18. The method of any one of claims 1 to 17, wherein the method comprises performing at least(a).

19. The method of any one of claims 1 to 18, wherein the method comprises performing at least(b).

20. The method of any one of claims 1 to 19, wherein the method comprises performing at least(c).21 . The method of any one of claims 1 to 20, wherein the method comprises performing at least(d).

22. The method of any one of claims 1 to 21 , wherein the method comprises performing at least(e).

23. The method of any one of claims 1 to 22, wherein the antibody mixture is a polyclonal antibody mixture or a mixture of monoclonal antibodies.

24. The method of any one of claims 1 to 23, wherein the MS is liquid chromatography-MS (LC- MS) and the MS / MS is tandem MS / MS.

25. The method of any one of claims 1 to 24, further comprising expressing one or more recombinant antibodies or antibody fragments corresponding to the one or more antibodies identified by the method defined in any one of claims 1 to 14.

26. The method of claim 25, further comprising assessing the binding of the recombinant antibody or antibody fragment to a target antigen.

27. The method of claim 26, wherein assessing the binding of the antibody or antibody fragment comprises performing an immunoassay.

28. The method of claim 27, wherein the immunoassay is enzyme-linked immunosorbent assay (ELISA).

Images