Sample collectionThe blood specimens of three healthy donors were obtained from Discovery Life Sciences. The collection and research use of these specimens were approved by the WCG Clinical IRB. The donors provided informed consent for their participation and the research use of their biological samples. The donors were compensated.Detailed anonymized information about all donors is available in Supplementary Table 1. Gender information was not used in the study design. Blood and plasma were collected from all 3 donors, but this study focused on donor Subject 2, a 49-year-old female who received the third Moderna shot. Blood for analysis was drawn 60 days later, using BD Vacutainer CPTTM mononuclear cell preparation tubes and stored in RNAprotect® Cell Reagent following the supplier’s instructions. For proteomics analysis, blood was centrifuged at 2,000 × g for 15 min at 4 °C, and samples were aliquoted into 3 ml Matrix cryo vials, totaling 3 ml serum per patient. Additional samples were acquired to benchmark the methods presented in this manuscript. Adalimumab, Bevacizumab, Cetuximab, Rituximab, and Trastuzumab were purchased from Sino Biologicals and combined to create a simple controlled oligoclonal mixture.RNA extractionPBMCs from donor Subject 2 were thawed, centrifuged, and lysed with Buffer RLT (Qiagen) and beta-mercaptoethanol. The lysate was transferred to a QIAshredder spin column and centrifuged for 2 mins at 19,500 × g. The homogenized lysate was combined with an equal volume of 70% ethanol and proceeded to RNA extraction with the RNeasy Mini Kit (Qiagen) for RNA extraction. The extracted RNA concentration was measured with the RNA BR assay for Qubit fluorometry (Thermo-Fisher), and RNA integrity was assessed by the Agilent TapeStation 4150 (Agilent Technologies), resulting in an RNA Integrity Number (RIN) of 9.6.cDNA synthesis, BCR amplification, and sequencing library generationThe SMARTer Human BCR IgG IgM H/K/L Profiling Kit (TakaraBio) was used to create NGS libraries. Following kit instructions, one µg of RNA was used in the first-strand cDNA synthesis. Per the user manual, PCR cycles were increased from 16 to 18 cycles for PCR2. Library concentration was determined with the 1× dsDNA HS assay for Qubit fluorometry (Thermo-Fisher), and sizes were assessed using the D1000 ScreenTape on the Agilent TapeStation 4150 (Agilent Technologies), targeting 640 bp for light chains and 690 bp for heavy chains. Peaks at 350 bp indicated adapter dimers and libraries with such peaks underwent size selection with NEBNext Sample Purification Beads (0.65× beads to sample volume) per the SPRIselect User Guide.Next-generation sequencingThe final products were re-quantified with the 1× dsDNA HS assay (Thermo-Fisher), normalized, and pooled to a 4 nM library. Following the Illumina Denature and Dilute Libraries Guide, an 8 pM loading library was prepared. To ensure complete denaturation and efficient binding to flow cells, the libraries were heat-shocked at 96 °C for 2 min and then cooled in an ice water bath for 5 mins. Sequencing was performed on an Illumina MiSeq platform using the reagent kit v3 for 600 cycle paired-end reads, including a 30% PhiX v3 spike-in.IgG enrichment against the antigenTotal IgG was enriched from 3 mL human serum using 3 mL of settled protein G agarose resin (Genscript) in a 20 mL gravity flow column (Biorad). The column was equilibrated with two 15 mL washes of 10 mM phosphate buffered saline (PBS). The serum was centrifuged at 23,000 × g for 10 min at 4 °C, combined with 9 mL of 10 mM PBS, and passed twice over the protein G resin column by gravity flow. After three 10 mL PBS washes, the IgG fraction was eluted with 12.5 mL of 0.1 M glycine buffer (pH 2.5). Eluted IgG was concentrated and buffer-exchanged into 10 mM PBS using a 30 kDa Amicon filter (Sigma), yielding 22.2 mg of total IgG. Anti-RBD antibody enrichment involved biotinylation of 0.4 mg of SARS-CoV-2 spike protein RBD (Exonbio), followed by coupling to 54 µL streptavidin-coated Sepharose beads (Cytiva) for 1 h at 4 °C. After incubation with total IgG, 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 glycine pH 2.5. Beads were washed with 500 µL 10 mM PBS to bring pH to 7, then 1 h UV elution was performed in 100 µLPBS using UV Stratalinker 2400 (Stratalinker), with 365 nm bulb to elute any remaining anti-RBD antibodies (less than 10% of the total polyclonal were eluted that way). A total of 190 µg anti-RBD antibodies were collected.In-solution digestionThe in-solution digestion process was performed on the antigen-enriched fraction described above (25 µg). Initially, the sample was concentrated to 100 µL via vacuum centrifugation in a low-pressure centrifuge (i.e., Speedvac). Following this, the sample underwent reduction using dithiothreitol (DTT) at a final concentration of 30 mM for 15 min at 95 °C. Subsequently, the sample was divided into two tubes for further treatment. In the first fraction (5/7th of the sample), alkylation with Iodoacetamide (IAA) was carried out at a final concentration of 50 mM for 30 min in darkness at room temperature. After the IAA treatment, the sample was precipitated using three volumes of acetone, followed by incubation at −20 °C, centrifugation at 23,000 × g for 10 mins at 4 °C, and drying of the pellet in a low-pressure centrifuge. The pellet was reconstituted with 10 µL of 4 M urea and incubated at 37 °C for 10 min. Subsequently, 90 µL of water was added, and the sample was divided into five tubes, with four tubes receiving 30 µL of 50 mM ammonium bicarbonate for overnight digestion with Trypsin, LysC, AspN, and Chymotrypsin at a 1:20 ratio. In the fifth tube, Pepsin digestion was performed by adding 30 µL of HPLC-grade water with 2 µL of 1 N HCl and Pepsin at a 1:20 ratio, followed by digestion at 37 °C for 15 min and inactivation at 95 °C for 3 min. Meanwhile, the remaining fraction (2/7th) was treated with 0.5 M 2-bromoethylamine hydrobromide (BEA) in a 100 mM tris buffer at pH 8 for 4 h at 25 °C. BEA converted cysteine into a lysine-like amino acid, making it a potential protease cleavage site for both Trypsin and LysC29. The BEA-labeled samples were precipitated by adding trichloroacetic acid to a final concentration of 20% (w/v). The precipitated sample was washed with acetone, dried, reconstituted, and digested with Trypsin and LysC overnight at 1:20 in 30 mM ammonium bicarbonate. All protease digests were dried under low pressure, reconstituted in 40 µL of 0.1% formic acid, and loaded onto Evotips in a method similar to Bache et al.41 and analyzed using HCD mode on Evosep-Exploris 240 (Thermo-Fisher). The proteases utilized in the polyclonal digestion were sourced from Promega.For non-reduction in solution digestion, 20 µg of polyclonal protein was reconstituted in 20 µL of 8 M urea in 100 mM tris buffer at pH 6.5. N-ethylmaleimide (NEM) was added to a final concentration of 2 mM and incubated at 37 °C for 2 h. Endoprotease LysC was added at a 1:20 protease to protein ratio and incubated overnight at 37 °C, followed by 4-h digestion with AspN at the same ratio and conditions. After protease digestion, the sample was dried under low pressure, reconstituted in 40 µL of 0.1% formic acid, and loaded with 2.5 µg of the digested samples onto Evotips for analysis using a Thermo Orbitrap Exploris 240 mass spectrometer, as detailed in the Mass spectrometry analysis section.Gel-based separationGel-based separation was conducted using Native and “Non-reducing Room Temperature” (NRT) gels. Native gel: polyclonal antibodies were separated on a Biorad precast 7.5% polyacrylamide Mini-Protean TGX gel with and without IdeS treatment, respectively. IdeS treatment involved incubating 50 µg of the polyclonal antibody with 50 units of IdeS (Promega), and the volume was reduced to 30 µL using vacuum centrifugation under low pressure. For the non-IdeS sample, 25 µg was loaded onto the gel. NativePAGE 4× buffer (Thermo-Fisher) was added in a 4:1 ratio for both IdeS-treated and non-treated samples, and the gel was run using a Biorad power unit set to 130 V for 180 min. The running buffer was diluted from 10× stock Tris/Glycine Buffer (Biorad) to 1× using Milli-Q water. Non-reducing Room Temperature gel (NRT, as shown in Supplementary Fig. 1A): The polyclonal antibody underwent separation in a modified non-reducing SDS-PAGE procedure without pre-heating, enabling different denaturation and migration patterns. A total of 10 µg sample per lane was combined with Laemmli buffer (Biorad), loaded onto a precast 7.5% Mini-PROTEAN TGX gel, and run at 150 V for about 60 min in Tris/Glycine/SDS electrophoresis buffer. All gels and buffers were sourced from Biorad. Coomassie brilliant blue (Biorad) staining for 30 min and overnight destaining with Biorad Destain were performed for all gels.In-gel digestionThe procedure used was a modified version of Mann’s method42. After gel separation, bands were cut and washed twice with 200 µL of HPLC-grade water. The bands were dehydrated with 200 µL of 100 mM tetraethylammonium bicarbonate (TEAB) and acetonitrile (ACN) (1:1 ratio). Subsequently, they were reconstituted in 25 mM DTT in 100 mM TEAB; samples were reduced at 56 °C for 30 min. The DTT solution was removed, and 55 mM IAA was added to alkylate for 30 min at room temperature. After that, the bands were washed twice with 0.4 mL of HPLC-grade water and dehydrated with 200 µL of 100 mM TEAB/ACN in a 1:1 ratio. For digestion, trypsin was diluted to a concentration of 6 ng/µL in 100 mM TEAB, of which 100 µL was added to gel pieces. The digestions were allowed to proceed at 37 °C overnight. The supernatant containing the digested peptides was collected in fresh tubes, and the gel pieces were dehydrated again to extract additional peptides using 100 µL of a 60% ACN, 0.1% formic acid (FA) solution. The supernatant from this extraction was combined with the digestion supernatant and dried down under low pressure. The dried samples were then resuspended in 40 µL of 0.1% FA, and 100% of the sample was loaded on Evotips according to the manufacturer’s instructions. Finally, the samples were analyzed in High Collision Dissociation (HCD) mode on an Evosep-Exploris 240 mass spectrometer, as detailed in the Mass spectrometry analysis section.Isobaric ambiguity resolution by EThcD mass spectrometryTo resolve Isoleucine/Leucine isobaric ambiguity, three selected samples underwent EThcD mode analysis as Zhokhov et al.31 proposed. These samples included in-solution digested trypsin and pepsin, with, in addition, the pepsin sample labeled at the C-terminal with arginine methyl ester dihydrochloride in a procedure similar to the one presented by Krusemark et al.43. For the labeling process, the coupling reagent 7-Azabenzotriazol-1-yloxy)tripyrrolidinophosphonium hexafluorophosphate, PyAOP, was utilized along with arginine methyl ester dihydrochloride. After resuspending the digestions in 10 µL of DMSO, 6 µL of a coupling solution was added (66 mg PyAOP into 132 µL DMSO) and incubated for 5 min. Following this, 14 µL of a Reagent Solution, consisting of 100 mg of Methyl Arginine dissolved in 50 µL of ddH2O plus 26 µL of N-Methyl morpholine, was added and incubated for 1 h. The reaction was then stopped by adding 320 µL of 0.1% formic acid and further subjected to liquid-liquid phase separation with 640 µL of chloroform for clean-up. The processed sample (i.e., the water phase) was dried and resuspended in 0.1% FA, and 2.5 µg of the digested samples were subsequently loaded onto Evotips according to the manufacturer’s instructions and then analyzed on an Evosep-Eclipse in EThcD mode as detailed in the next section.Mass spectrometry analysisMass spectrometry analysis was conducted using an Orbitrap Exploris 240 (Thermo-Fisher) for HCD-MS experiments and a Thermo-Fisher Orbitrap Eclipse Tribid Mass spectrometer for EThcD-MS experiments. The HCD-MS experiments involved 30 samples per day on a 15 cm PepSep column from Bruker Daltonics, with data acquired at a resolution of 60,000 for 400–2000 m/z precursors. A standard AGC target and maximum injection time set to Auto were used, along with an intensity threshold of 2.5e4 and charge states 2–8 included. Dynamic exclusion was set to 15 s, and MS/MS fragmentation with fixed 30% HCD at a resolution of 7500. On the other hand, EThcD-MS experiments were carried out on an Orbitrap Eclipse Tribid Mass spectrometer connected to an Evosep with similar settings as the HCD-MS experiments but using ETD for fragmentation with EthcD settings and fixed energy at 55. MS/MS spectra were acquired at 7500 resolution with a maximum injection time of 50 ms.Recombinant expressionFab Sequences were sent to Sino Biologicals for protein expression using human IgG1 Fc as a backbone. The target gene was amplified, inserted into an expression vector, confirmed through sequencing, and passed on to downstream processes. Expression was performed using HEK293 cells for transient transfection in a serum-free medium. After 6 days of culture, proteins were purified using an affinity purification protein G column and analyzed by SDS-PAGE.Surface plasmon resonance, SPRAffinity was carried out with RBD immobilized on the OpenSPR-XT instrument (Nicoya Lifesciences). RBD was immobilized in the active channel using the standard EDC/NHS amine coupling chemistry for amine coupling to high-capacity carboxyl sensors and 1 M Tris-based solution for reaction quenching (Supplementary Fig. 2A). The immobilization setup provided sufficient RBD levels (3150 RU). In contrast, PBS with 0.05% Tween20 was used as the running buffer to minimize non-specific binding (Supplementary Fig. 2B). A regeneration screen identified 10 mM Glycine-HCl pH 1.5 as a universal regeneration solution for subsequent analysis. Screening of a nanomolar concentration range for antibody analytes identified 100 nM as the optimal concentration for evaluating all candidates. Normalized sensorgrams for each concentration set were overlaid for comparative analysis (Supplementary Fig. 2C), and measurement of complex stability was plotted as immediately post-injection (early stability) vs 30 s after (late stability). The same experimental setup performed an off-rate screen with 300 s contact time and 600-s dissociation of 100 nM analytes normalized for their relative binding responses. It ranked based on relative dissociation percentages (Supplementary Fig. 2D).ELISAIndirect ELISAs were performed on patient plasma and antibodies (i.e., PD124 polyclonal and mAbs sequenced from PD124), respectively. For the ELISA performed on plasma (Fig. 2A), 0.1 µg of recombinant RBD (ExonBio) was immobilized on Maxisorp 96-well plates (Thermo-Fischer) overnight at 4 °C (2 µg/mL in sodium carbonate-bicarbonate buffer pH 9) whereas 0.2 µg of recombinant RBD was immobilized to plate for the recombinant mAb ELISA (Fig. 4A). For the negative controls in ELISA testing of the recombinant antibodies, we used human IgG antibodies purchased from Sigma Aldrich (I4506) (Fig. 4A–C). Human plasma was obtained from Innovative Research Inc (IPLASK2E50ML, single donor human plasma with K2 EDTA) for testing the titer shown in Fig. 2A. For the positive controls, we used the following antibodies: Anti-SARS-CoV-2 Spike RBD Neutralizing Antibody, Human IgG1 AS35 (Acro Biosystems, clone AS35, lot no. S35-21CHF1-ZS) (see Fig. 4B, C), and Anti-Spike RBD Neutralizing Antibody, S1N-M122 (Clone AM122) from Acro Biosystems (see Fig. 4D). Uncoupled antigen was removed by washing wells three times with 200 µL of 0.03% Tween in PBS (PBS-T), followed by blocking with SuperBlock buffer (Thermo-Fischer). Plasma or recombinant mAbs were serially diluted with Stabilzyme diluent, then 100 µL was added to wells in duplicate and incubated for 1 h at room temperature. Uncoupled antibodies were removed by three washes with 200 µL PBS-T. Goat anti-human IgG Fc−horseradish peroxidase-conjugated secondary antibody (Sigma Aldrich, catalog no. A0170, lot no. 0000181901) was diluted 1:5000, then 100 µL was added per well and incubated for 1 h at room temperature. The unbounded secondary antibody was removed by washing three times with PBS-T, then 100 µL 3,3′, 5,5′ tetramethylbenzidine (TMB) substrate (Thermo-Fischer) was added and allowed to react for 3 min, followed by quenching with 100 µL 0.2 N sulfuric acid (Sigma Aldrich) to stop the reaction. The plates were read in a spectrophotometer (Biotek Synergy LX) at 450 nm, and the average OD450 was plotted for each sample to generate the sigmoid binding curve.ACE2 competition binding assayAs described above, 0.2 µg of recombinant RBD (ExonBio) was immobilized on Maxisorp 96-well plates (Thermo-Fischer) overnight at 4 °C. After washing and blocking, wells were incubated with serially diluted recombinant antibodies. Positive and negative controls included Sars-CoV2 Spike RBD neutralizing antibody AS35 (Acro Biosystems, clone AS35, lot no. S35-21CHF1-ZS) and human IgG from serum (Sigma Aldrich), respectively. Biotinylated ACE2 (Sino Biologicals) (70 ng/mL) and streptavidin-HRP conjugate (Millipore) (16.67 ng/mL) were added sequentially. TMB substrate (Thermo-Fisher) was used, followed by quenching with 0.2 N sulfuric acid (Sigma Aldrich). Plates were read at 450 nm on a Biotek Synergy LX spectrophotometer to generate the neutralization curve based on the average OD450 for each sample. The positive control is AS35, an Anti-SARS-CoV-2 Spike RBD Neutralizing Antibody, Human IgG1.In vitro, cell-based neutralization assayFor the in vitro, cell-based neutralization assay, HEK293/Human ACE2 Stable Cell Line (Acro Biosystems) was cultured in complete DMEM medium (Shanghai BasalMedia Technologies) supplemented with 10% fetal bovine serum (VivaCell) at 37 °C with 5% CO2. Recombinant antibody samples were serially diluted in a complete DMEM medium and added to a 96-well plate. Pseudovirus SARS-CoV-2 Spike (WT) (PSSW-HLGB001, Acro Biosystems) was diluted and added to the plate. For the cell control group, 25 μL of the complete DMEM medium was added instead. After 60 min of incubation, cell density was adjusted, and cells were seeded into the plate for a 48-h incubation. A detection reagent (Britelite plus Reporter Gene Assay System) was added to each well. The positive control is S1N-M122, a known anti-Spike RBD Neutralizing Antibody, Chimeric mAb, Human IgG1 (Clone AM122), sourced from Acro Biosystems.Luminescence measurementThe luminescence value of each well in the plate was measured using a microplate reader. The detection time for each well was set to 0.1 s./well.$${{\rm{Inhibition\; rate}}}=\left(1-\frac{X\,-\,\overline{{{\rm{CC}}}}}{\overline{{{\rm{VC}}}}\,-\,\overline{{{\rm{CC}}}}}\right)\times 100\%$$X, the luminescence value (RLU) of a given well.CC, cell control, only adds cells.VC, virus control, and only cells and pseudovirus are added.\(\overline{{{\rm{CC}}}}\), the mean value of the cell control group.\(\overline{{{\rm{VC}}}}\), the mean value of the virus control group.The data were analyzed using GraphPad Prism 8 software (Nonlinear regression). IC50 or relative IC50 is defined as the incubation concentration of a sample required to bring the curve down to the point halfway between the top and bottom plateaus of the curve. Absolute IC50 is defined as the incubation concentration of a sample that provokes 50% inhibition.Data analysisIgSeq data processingThe protein database for searching the MS data was generated by translating the merged R1 and R2 reads from the Illumina MiSeq run using the human IMGT germline database as a reference. After translation, identical protein sequences were merged into unique sequences.MS-based database searchingNovor.Cloud (http://novor.cloud) was used to match the mass spectrometry data to the antibody sequence database that is translated from the NGS reads. The matching antibody sequences were sorted by their MS coverage in CDR regions and total sequence coverage, and further organized by clustering based on their CDR3 sequences.De novo protein sequencingDe novo antibody protein sequencing is performed at three different scale levels: (1) Contig assembly, (2) within-chain CDR pairing/chain assembly, and (3) heavy-light chain pairing.Scale 1, contig assemblyThe contig assembly consists of five steps. Step 1 – de novo peptide sequencing: The MS/MS spectra were de novo sequenced using Novor software to obtain peptide sequences44. Additionally, Novor assigns a confidence score for each amino acid within a given sequence. Step 2 – finding the maximum mass block overlap: The algorithm computes the maximum “mass block” overlap between every possible pair of peptides. This involves considering amino acid permutations within a block (e.g., [DF] = [FD]) and isobaric similarities (e.g., [I] = [L], [N] = [GG], [Q] = [GA] = [AG]) and direct similarity [A] = [A]. This concept is best illustrated with an example. Suppose LGAGYDFNH and FDGGHWYQQL are two peptides. The underlined sub-sequences can be divided into blocks [DF][N][H] and [FD][GG][H], respectively. Since mass [DF] = mass [FD], mass [N] = mass [GG], and mass [H] = mass [H], the two peptides overlap by 3 mass blocks. The two overlapping peptides can be merged to form a longer sequence LGAGY\({\beta }_{1}{\beta }_{2}\)HWYQQL where \({\beta }_{1}=\) DF or FD, and \({\beta }_{2}=\) N or GG are the mass blocks being considered. The one with the highest average Novor amino acid confidence score is used between the two sequence choices of each block. Step 3 – overlap graph construction: To construct the overlap graph, each de novo peptide with a Novor score of 75 or higher corresponds to a vertex/node. An edge is added between each pair of vertices if their peptides overlap by at least 3 mass blocks. Step 4 – contig construction: For each path consisting of \(k\) (\(k\le 3\)) vertices in the graph, a contig is constructed by merging the peptides of all vertices on the path. Each contig is then aligned with its best matching germline gene sequence. Contigs that completely cover a CDR region (including two additional amino acids on both sides of the CDR) are retained. The Novor.Cloud engine (available online at http://novor.cloud) was used to search the MS/MS spectra in the contig database. The contigs were grouped by CDR region sequences and sorted according to their database search scores. Step 5 – manual human inspection: The grouped contigs were subjected to human inspection to select a short list of CDR region sequences and their corresponding contigs. Generally, longer contigs, with more PSM coverage and longer peptide overlaps, were given higher priority. Subtle de novo peptide sequencing errors were also corrected during human inspection.Scale 2, CDR pairing and chain assemblyThe assembled and selected CDR contigs were subjected to pairing analysis. This process is analogous to the heavy-light chain pairing described below, but with contigs replacing the chains. After computing the similarity scores for all pairs of CDRs, pairs that showed high similarity scores to each other, but relatively low similarity scores to other CDRs, were selected and regarded as being from the same antibody chain. Clusters of high-scoring CDR1, CDR2, and CDR3 pairs were assembled by aligning them to a germline gene. Any gaps in the framework region were filled with either overlapping de novo peptides or germline gene sequences to obtain a full chain sequence. Additionally, long peptides spanning two CDR regions were identified by database searching, and disulfide bridge-containing peptides from non-reduced protease digestion were identified using pLink software45. These peptides crossing two CDR regions were utilized to help resolve ambiguities in the quantification-based pairing.Scale 3, heavy-light chain pairingNovor.Cloud was used to search the MS/MS data for the heavy and light chain sequences to identify PSMs. MaxQuant software (Ver2.1.3) calculated the peak area of each peptide from each fraction in NRT and native gel separation experiments. For a given peptide with peak areas \({y}_{1},{y}_{2},\ldots,\,{y}_{k}\) in the \(k\) fractions of a separation experiment and \(\alpha={\sum }_{i=1}^{n}{y}_{i}\), the normalized quantification vector of the peptide is calculated by \(y=\left(\frac{{y}_{1}}{\alpha },\frac{{y}_{2}}{\alpha },\,\ldots,\frac{{y}_{k}}{\alpha }\right)\). The co-existence of two different peptides being part of the same antibody was determined using the Pearson correlation coefficient between their normalized quantification vectors. The similarity score between two chains was computed as the average similarity between every pair of unique peptides from the two chains. In cases of multiple separation experiments, the similarity scores from each experiment were summed to obtain the final similarity score. After computing the scores for all pairs of heavy and light chains, pairs with high similarity scores to each other and relatively low similarity scores to other chains were selected and regarded as originating from the same antibody. A Heavy-Light chain pairing matrix can be found in Supplementary Fig. 1B.Isobaric ambiguity resolutionTo resolve isobaric ambiguities between leucine and isoleucine using Electron Transfer/Higher-energy Collisional Dissociation (EThcD) mode, z-ions were identified via Novor de novo protein sequencing or database matching. Once isobaric ambiguity was confirmed, specific w-ions indicative of leucine or isoleucine were manually assessed by examining the presence or absence of w-ion peaks unique to either residue. Specific w-ion peaks confirmed the identity of leucine or isoleucine at distinct positions, thereby resolving the isobaric ambiguity in the sequence.Non-reduced samples digestion analysisThe identification of the S–S bridge-containing peptides from non-reduced protease digestion was analyzed using pLink45. The software identification parameters were set as follows: Flow Type: Disulfide bond (HCD-SS); Enzyme: LysC-AspN; Peptide mass: 300-9000 Da; 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 (Supplementary Table 2) contained the sequences of interest deduced from the polyclonal antibody analysis.Statistics & reproducibilityNo statistical methods were employed to predetermine sample size, and no data were excluded from the analyses. The experiments were not randomized, and investigators were not blinded to allocation during the experiments or outcome assessment. Due to limited availability of material, experiments using the serum pAb from Subject 2 were carried out only once. A benchmark experiment with a mixture of five mAbs was used to demonstrate the reproducibility of similar results.Reporting summaryFurther information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
