key: cord-341701-zropd3mo authors: Adhikari, Subash; Nice, Edouard C.; Deutsch, Eric W.; Lane, Lydie; Omenn, Gilbert S.; Pennington, Stephen R.; Paik, Young-Ki; Overall, Christopher M.; Corrales, Fernando J.; Cristea, Ileana M.; Van Eyk, Jennifer E.; Uhlén, Mathias; Lindskog, Cecilia; Chan, Daniel W.; Bairoch, Amos; Waddington, James C.; Justice, Joshua L.; LaBaer, Joshua; Rodriguez, Henry; He, Fuchu; Kostrzewa, Markus; Ping, Peipei; Gundry, Rebekah L.; Stewart, Peter; Srivastava, Sanjeeva; Srivastava, Sudhir; Nogueira, Fabio C. S.; Domont, Gilberto B.; Vandenbrouck, Yves; Lam, Maggie P. Y.; Wennersten, Sara; Vizcaino, Juan Antonio; Wilkins, Marc; Schwenk, Jochen M.; Lundberg, Emma; Bandeira, Nuno; Marko-Varga, Gyorgy; Weintraub, Susan T.; Pineau, Charles; Kusebauch, Ulrike; Moritz, Robert L.; Ahn, Seong Beom; Palmblad, Magnus; Snyder, Michael P.; Aebersold, Ruedi; Baker, Mark S. title: A high-stringency blueprint of the human proteome date: 2020-10-16 journal: Nat Commun DOI: 10.1038/s41467-020-19045-9 sha: doc_id: 341701 cord_uid: zropd3mo The Human Proteome Organization (HUPO) launched the Human Proteome Project (HPP) in 2010, creating an international framework for global collaboration, data sharing, quality assurance and enhancing accurate annotation of the genome-encoded proteome. During the subsequent decade, the HPP established collaborations, developed guidelines and metrics, and undertook reanalysis of previously deposited community data, continuously increasing the coverage of the human proteome. On the occasion of the HPP’s tenth anniversary, we here report a 90.4% complete high-stringency human proteome blueprint. This knowledge is essential for discerning molecular processes in health and disease, as we demonstrate by highlighting potential roles the human proteome plays in our understanding, diagnosis and treatment of cancers, cardiovascular and infectious diseases. a. Establishing agreed, stringent, reliable standards, b. Identifying >1 protein product from each proteincoding gene and c. Detecting expression of the remaining missing proteins (see below). 2. To make proteomics an integrated component of multiomics studies to advance life sciences, biomedical sciences and precision medicine. Comparisons with the HGP are numerous. Both global projects are ambitious cooperative community efforts seeking to identify how genes (HGP) or proteins (HPP) help define the molecular mechanisms underlying health and disease. Both groups have implemented exhaustive data sharing and stringent quality control efforts. However, we now know that sequencing human genomes is necessary but not sufficient to understand the complexity of human biology or pathology. Knowledge of expressed proteins (including concentration, spatio-temporal localization, activities, protease-processed forms, transport, interactions, splice isoforms, PTMs and the many proteoforms derived from the proteome) cannot be predicted by genome sequencing alone. HPP structure and achievements HUPO launched the HPP in 2010 at the 9th HUPO Annual World Congress in Sydney, with Gil Omenn as the inaugural chair. The HPP started without long-term funding. Financial support over the decade came through individual principal investigator projects, institutional infrastructure/core facilities and philanthropy, without large-scale, long-term, integrated multi-government strategic funding. The HPP grew from several archetypal HUPO projects (plasma, liver, brain, cardiovascular, kidney/urine proteomes). At present, the HPP comprises two strategic initiatives-chromosome-centric (C-HPP; 25 teams) and biology/disease-centric (B/ D-HPP; 19 teams)-organized in a strategic matrix underpinned by four Resource Pillars: antibodies (AB), mass spectrometry (MS), KB and Pathology (Fig. 1a) . Before explaining these elements in more detail, we will describe HPP's criteria underlying the establishment of the high-stringency human proteome. Defining the human proteome at high-stringency. The HPP relies upon a coordinated system developed by neXtProt and UniProtKB, which attributes five levels of supporting evidence for protein existence (PE) 6 . Evidence at PE1 indicates clear experimental evidence for the existence of at least one proteoform, based on credible identification by MS, Edman sequencing, Xray, nuclear magnetic resonance (NMR) structure of purified natural protein, reliable protein-protein interaction and/or antibody data. PE2 indicates evidence limited to the corresponding transcript (cDNA, reverse-transcriptase PCR, northern blotting). PE3 indicates the existence of orthologs in closely related species. PE4 refers to entries based on gene models without evidence at protein, transcript or homology levels. PE5 classifications indicate that coding evidence is doubtful and/or probably corresponds to an incorrect in silico translation of a non-coding element. As PE5 entries are largely non-coding, the HPP preferentially tracks only PE1,2,3,4 protein-coding entries. Proteins classified as PE2,3,4 are colloquially referred to as missing proteins 7, 8 . Since 2013, neXtProt and the HPP (Fig. 1b) have issued annual tallies of PE1,2,3,4,5 status (www.nextprot.org/about/protein-existence), which have been reported in annual collaborative metrics manuscripts (ref. 9 and references therein). The latest neXtProt HPP reference release (https://www. nextprot.org/about/statistics, data release: 17-01-2020) designates credible PE1 evidence for 90.4% of the human proteome (17, 874 PE1s from the 19,773 PE1,2,3,4 protein entries excluding the dubious 577 PE5 entries). This leaves 1899 (9.6%) PE2,3,4 missing human proteome entries to be identified at highstringency. Here, the term high-stringency refers to rigorous HPP standards for post-acquisition processing and any protein inference made from raw MS peptide spectral data 10 . This term avoids confusion with the pre-existing term 'high-accuracy' Box 1 | HPP decadal achievements • Generated a framework, plan and governance structure for community-based mapping of the human proteome. • Confirmed neXtProt as the HPP reference knowledgebase and supported creation and use of ProteomXchange (PX) to register and make proteomics raw mass spectrometry (MS) and metadata available and reusable under FAIR (findable, accessible, interoperable and reusable) principles. • Engaged neXtProt, PeptideAtlas, PRIDE and MassIVE as partners in the generation of annual neXtProt HPP release and a high-stringency HPP knowledgebase (KB). • Encouraged community support of high-stringency protein inference and proteomic data analysis. • Built MS data interpretation guidelines that promoted the application of standardized analysis of community human MS and proteomic data to progressively complete the human proteome parts list. • Aligned with the Human Protein Atlasʼ (HPA) cell and tissue spatiotemporal maps in health/disease and supported community efforts to raise awareness of antibody specificity and quality assurance issues. • Partnered with SRMAtlas to develop quantitative targeted proteomics assays for the analysis of key proteins and hallmark pathways/networks. Proposed and built global collaborative initiatives to investigate the biology of human health and disease at a proteomic-wide scale. • Raised the profile and visibility of proteomics, as an essential component of life sciences and biomedical research by promoting the development of instrumentation and methods for proteoform analysis, as well as activity/function that cannot be addressed by genomics. • Initiated a programme to determine the biological function for uncharacterized PE1 proteins (see below) that currently lack functional annotation. • Established a HUPO Early Career Researcher network to engage, mentor and highlight research from young scientists/clinicians, while actively promoting gender and regional balance. • Establish a community initiative to systematically map all human proteoforms. • Establish optimized workflows for human proteome detection, quantification and functional characterization, including low abundance and/or temporo-spatially restricted proteins. • Continue to support and promote the provision of technical standards, metrics and stringent guidelines for confident protein identification and quantification. • Create a comprehensive, accurate, publicly-accessible, reference human proteome knowledgebase, reusable under FAIR principles. • Maintain education and training programmes in all aspects of proteomics including proteomic data analysis for early career researchers and clinical scientists. • Be a focal point for life sciences researchers, pathologists, clinicians and industry communities seeking to translate and leverage proteomic and proteogenomic data to improve human health through: (i) greater understanding of the molecular mechanisms of common and rare diseases, (ii) identification of pathophysiological changes to generate disease and wellness diagnostic biomarkers, and (iii) development of new effective and safe personalized therapeutics. frequently used in MS parlance, which relates to the generation of high (mass)-accuracy spectra from modern instruments. The use of high-stringency influences the quality of all protein inferences derived from any raw MS data. The HPP routinely applies high-stringency protein inference analytics 10 , generated through the Trans-Proteomic Pipeline 11 . Claims for detection of new PE1s and/or detection of coding elements not previously in neXtProt, should meet minimum evidence thresholds. The current HPP guidelines 10 require at least two uniquely mapping peptides (using neXtProt's Peptide Uniqueness Checker Tool 12 ), which are at least nine amino acid residues long. Peptides must be non-nested (i.e., one not fully contained within another) but may overlap partially so coverage exceeds >18 residues. The HPP also requires full declaration of false discovery rate (FDR) calculation procedures at peptide and protein levels, with a maximum allowable protein-level FDR of 1%. Within HPP missing protein publications, further validation is required through synthetic peptide spectra matching 13 . Going forward, the HPP also requests association of Universal Spectrum Identifiers with every MS spectrum 10 . We emphasize that MS-based data from high-accuracy 14,15 MS instruments combined with subsequent high-stringency protein inference analysis provide definitive best-practice confirmation of protein identification and abundance. To ensure high-quality analyses, increasingly stringent criteria have been applied 10 spectral data Findable, Accessible, Interoperable and Reusable (FAIR) 17 . Many previous studies use high (mass)-accuracy instruments with subsequent protein inference identifications undertaken at lower default settings, such as accepting single peptides or those only seven amino acids in length and/or not conforming to more rigid neXtProt proteotypic analysis 18, 19 . These analyses can result in spurious identification of many more false positives 20 , with lower-quality single non-proteotypic spectra data colloquially referred to as one-hit wonders, better explained by sequence variation in other highly observed proteins 21 . Chromosome-centric (C)-HPP. The C-HPP (https://www.hupo. org/C-HPP) aims to annotate all genome-encoded proteins 7, 8 in an unbiased and high-stringency manner. It explores proteins that have not previously been confidently observed by MS or other analytical methods 7, 8, 22 . International C-HPP teams are organized according to chromosomes (Chrs), namely Chrs 1-22, Chr X, Chr Y and mitochondrial (Mt) genome teams. From 2017, the C-HPP expanded its mission to include functional characterization of the 1899 PE2,3,4 proteins that have not been confidently observed and 1254 PE1s that have no neXtProt curated function (uncharacterized PE1s or uPE1s), cumulatively referred to as the dark proteome 23, 24 . Biology/disease-centric (B/D)-HPP. The B/D-HPP (https:// www.hupo.org/B/D-HPP) measures and interprets human proteome data under a range of physiological and pathological conditions. It focuses on the following: (i) elucidating the hallmark protein drivers of biology/disease and (ii) promoting development of new proteomics analytical tools such as Ab-based approaches and targeted selected/multiple/parallel reaction monitoring (SRM/MRM/PRM) assays. As an example, the initial HUPO liver proteome project grew into a B/D-HPP team focussing on liver expression profiles, PTMs, tissue expression, subcellular localization, interactions, physiology and pathologies 25 . The Chinese CN-HPP have characterized four liver cell types, emphasizing benefits of acquiring cell-type specific maps to understand underlying biology/pathology 26 . In addition, they mapped landscapes of early hepatocellular and lung carcinoma, generating cancer subtype alterations where proteomic signatures identified poor prognosis patients and/or those benefiting from targeted therapy 27, 28 . In other studies, they analysed microdissected cell types with gross anatomical resolution using MS 26, 29 , revealing circadian cycles and spatio-temporal proteome expression in the liver, brain, heart and stomach 30 , providing resources to better understand organ biochemistry, physiology and pathology. Significant discoveries continue to be made from all B/D-HPP teams across personalized cancer immunotherapy and therapeutic modalities (e.g., lymphoma 31 , ovarian 32 , liver 27 and lung 28 cancers), with PTMs orchestrating many outcomes including response to therapy [33] [34] [35] . B/D-HPP resources include the Human SRMAtlas 36 , a unique compendium of high-resolution spectra and multiplexed SRM/ MRM/PRM assays developed from 166,174 synthetic proteotypic peptides. This assay library enables targeted identification and quantification of a theoretical maximum of 99.7% of the human proteome 36 , provided proteins are expressed spatiotemporally at concentrations amenable to MS detection. For example, SRMAtlas supported the C-HPP Chr X team's confident identification of missing proteins 37 . HPP resource pillars. The B/D-HPP and C-HPP are supported by four HPP Resource Pillars that ensure effective data generation, integration and implementation, including the establishment of metrics and guidelines, enhancement of technology platforms, reagent development and optimal use of existing and emerging data streams (Fig. 1a) . The HPP MS Resource Pillar informs the community about MS technology/workflow advances, appropriate high-stringency standards and liaises with industry regarding instrument development, all leading to improved depth and accuracy of proteome identification, quantification and modification. These include methods like matrix-assisted laser desorption ionization time of flight (MALDI-TOF)-MS, electrospray-MS, bottom-up (shotgun) MS, data-dependent acquisition MS, data-independent acquisition (DIA) MS, targeted SRM/MRM/PRM, top-down MS, crosslinking MS, PTM analysis, N-and C-termini measurement, MS data computational analysis and interactomics. The MS Resource Pillar previously undertook a SWATH/DIA-MS reproducibility study 38 and are currently coordinating a phosphopeptide challenge involving >20 participating labs with partners SynPeptide Shanghai and Resyn Biosciences South Africa, who have provided a human phosphopeptide standard set with unphosphorylated counterparts, as peptide mixtures spiked into a yeast tryptic digest background. This will result in a better understanding of phosphopeptide enrichment, MS data analytics and informatics tools. The HPP Ab Resource Pillar, ostensibly led by the Human Protein Atlas (HPA; www.proteinatlas.org), was initiated in 2003 and uses Ab-based strategies to analyse spatio-temporal aspects of the proteome 39 . Linking the identification of proteins with 'realtime' localization at tissue, cell and subcellular levels supports a more comprehensive understanding of biology, health and disease. This requires information at resolution not currently available by MS (see single-cell section below). Approaches for spatio-temporal proteomics include single-cell in situ MS, fractionated cell lysates, proximity labelling or imaging-based proteomics 40, 41 . Imaging-based proteomics has a clear advantage, namely analysing proteins in their native location at single-cell resolution. To this end, the HPA has developed industrial scale epitope-directed Abs for community use. HPA also integrates multi-omics data. It contains extensive transcriptome data and neXtProt PE assignments, and contributes to the open-access catalogue Antibodypedia, containing >4 M Abs (www.antibodypedia.org) 42 against >19,000 targets that assist the community to select application-appropriate Abs. At HPP launch, the HPA had detected >50% of the proteincoding genome 43 . Currently,~87% of the proteome is targeted by >1 HPA Ab, detected though an encyclopaedia of >10 M annotated high-resolution digital images, partitioned into a number of sub-atlases that are interconnected [44] [45] [46] [47] [48] . These currently comprise the; Tissue Atlas (protein distribution across all major tissues), Cell Atlas (subcellular localization and heterogeneity in single cells), Pathology Atlas (correlations between gene expression and patient survival in major human cancer types), Blood Atlas (protein profiles across major immune cells and blood levels), Brain Atlas (protein distribution in the brain), and the Metabolic Atlas (various tissue metabolic enzymes localizations). Over the decade, HPA's open-access database [44] [45] [46] [47] [48] has become one of the world's most visited biological resources (>3.6 M visits in toto annually). The HPA also plays an emerging role in establishing guidelines around the appropriate use of Abs and ensuring immunoassay validation [49] [50] [51] . It recently spent considerable effort validating the selectivity of their Abs, including championing efforts of the International Working Group for Ab Validation proposing many new approaches implemented across >10,000 Abs 42 . In addition, HPA's massive collection of images has supported a multitude of publications and become a citizen science resource for developing AI classification learning models 52, 53 . The HPA is sustained PERSPECTIVE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-020-19045-9 through community contribution to ELIXIR 54 , that allows scientists from academia and industry to explore spatio-temporal aspects of the human proteome [55] [56] [57] . Since 2018, the HPP Pathology Resource Pillar has coordinated identification of areas of unmet clinical need, develops fit-forpurpose clinical assay guidelines/standards, promotes best-practice awareness, coordinates access to quality clinical samples/metadata and liaises with pathology organizations, diagnostic companies and regulatory agencies to promote professional application of proteomics in pathology. The HPP KB Resource Pillar captures, collects, collates, analyses and re-distributes all human proteome data. As cohesive knowledge transfer plays such a crucial role in big data science, including the HPP, the KB pillar's activities are addressed in the expanded section that follows. The human proteome in the neXtProt HPP reference KB Assembly and curation of neXtProt. Prior to the HPP, HUPO established a Protein Standards Initiative (PSI; www.hupo.org/ Proteomics-Standards-Initiative), emphasizing from the outset their priority for defining high-quality community standards, minimal requirements for experimental information 58 and highstringency data metrics. PSI continues to work cooperatively with the HPP KB to inform HPP initiatives, pillars and teams. In 2013, neXtProt 59,60 was officially designated as the HPP reference KB 61 . Annually, a neXtProt release is designated as the 'HPP release' and this serves as the basis for subsequent HPP high-stringency analyses, planning and reporting progress 10 . It receives and curates data from UniProtKB/SwissProt 62 , adding MS evidence from PeptideAtlas 63 and since 2019 from Mas-sIVE 64 . neXtProt also curates Ab-based, genome, transcriptome and other biological data to create an assembled snapshot of the human proteome 6, 65 . PeptideAtlas (http://www.peptideatlas.org) uses sequence search engines Comet 66 , X!Tandem 67 and SpectraST 68 to reprocess publicly uploaded MS/MS data deposited through ProteomeXchange (PX). Data are aggregated using rigorous criteria including peptide spectral matching with FDR~0.0009% in the latest PeptideAtlas build to achieve ≤1% FDR at the protein level. MassIVE searches public datasets using the MS-GF+ search engine 69 , also with strict criteria 70 to enforce global <1% FDR at the protein level for single-peptide identifications and stricter <0.01% FDR for proteins identified by >2 peptides. PeptideAtlas and MassIVE peptide lists are integrated into current neXtProt builds. neXtProt then cross-references all peptides to protein entries and validates PE levels, requiring at least two MSidentified uniquely mapping 9-mer non-nested peptides coming from either PeptideAtlas or MassIVE. neXtProt builds on UniProtKB/SwissProt PE1 entries that include MS, partial or complete Edman sequencing, X-ray or NMR structure, reliable protein-protein interaction data or Ab detection by considering additional PeptideAtlas/MassIVE data that meets minimum peptide uniqueness, number, length, nestedness and other requirements to upgrade entries to PE1. Noticeably, neXtProt is becoming more reliant on MS than non-MS data (e.g., 1860 non-MS data PE1s in 2016 down to 950 in 2020). To ensure only high-quality Ab data are used, in 2018 neXtProt/HPP Ab pillar revised criteria to upgrade entries to PE1, including specificity and other rigorous criteria suggested by the International Working Group for Ab Validation 50 . As an example, neXtProt recently analysed 41 Ab-based publications and upgraded three PE2,3,4 entries to PE1. Discussions regarding data provenance delivered by non-MS data for PE1 assignment actively continues in the HPP 71 . Expansion of the high-stringency proteome. The baseline number of protein-coding genes in the reference proteome given by neXtProt is managed by UniProtKB/SwissProt. Every proteincoding gene is assigned a protein entry (inclusive of all proteoforms), with chromosome location and other data organized under these entries. The number of protein-coding genes originally estimated by the HGP dropped from >100,000 72 to a relatively stable~19,700 at HGP draft release, a number that is close to the current neXtProt's 2020 release of 19,773 proteincoding genes ( Supplementary Fig. 1) 60 . Confident detection of the human proteome has consistently risen from 69.8% neXtProt PE1 entries in 2011 to 90.4% in 2020 (Fig. 2a) . However, progress has recently slowed (Fig. 2a) , suggesting that it may be difficult to confidently identify all 1899 missing proteins. Parenthetically, the HGP celebrated a provisional 90% completion ten years after its launch 5 and annotation of the human genome still remains incomplete or uncertain, especially in regions of repeat sequences and Z-DNA inserts. In each of the seven annual HPP metrics papers published to date ( 9 and refs therein), PE1,2,3,4,5 entries have been added, deleted, renamed, merged and/or de-merged, indicating ongoing fine-tuning of the HPP reference proteome. The Sankey flow diagram 73 (Fig. 2b) illustrates that significant fluidity has occurred across all PE classifications since 2011. The largest shifts occur with increases in PE1s (13,603 up to 17,874), followed by decreases in PE2s (5,587 down to 1,596), despite increasingly stringent HPP MS data guidelines adopted in 2015 (v2.1) and 2019 (v3.0). Linear regression of PE1 increases against PE2 decreases results in a strong (R 2 = 97%) inverse correlation, suggesting new PE1s come from PE2s where mRNA expression has been previously observed. Extrapolation of this PE1 discovery curve suggests that 95% PE1s may be reached sometime between 2024 and 2027. Minor conversions occurred between other PE categories, with downgrading of PE1 or PE2s and upgrading of PE4s, and even a few PE5s, to PE1s. Although a plethora of studies show low linear correlations (40-60%) between mRNA level and protein abundance 74,75 , our binary data (Fig. 2 ) supports the contention that once a neXtProt entry has mRNA expression verified, those PE2s are amenable to upgrade to PE1, whereas PE3s, PE4s and particularly PE5s are more resistant to PE1 upgrade. Missing protein analyses. neXtProt protein descriptors and associated data can be used to analyse protein groups/families according to their Chr location or PE classification. Missing protein (PE2,3,4) analysis indicates that some groups/families have been upgraded to PE1 more successfully than others (Fig. 3a) . For example, between 2011 and 2020, 372 zinc (Zn) finger proteins, 171 transmembrane proteins, 93 carbohydrate metabolism proteins, 90 testes-, sperm-, prostate-associated proteins, 78 coiled-coil domain-containing proteins and 58 homeobox proteins have been upgraded from PE2,3,4 to PE1. These represent the six most prominent protein groups upgraded to PE1 over the decade (Fig. 3a, green) . In contrast, two G protein-coupled receptor (GPCR) chemosensory families prove particularly resistant to PE1 upgrade: many olfactory receptors (ORs) are still missing in 2020 76 (417 of the 2011 PE2,3,4 entries, including putative and uncharacterized ones) and 17 of 20 taste receptors remain PE2,3,4 s (Fig. 3a, magenta) . In addition, some groups with a large number of PE1-upgraded protein entries still contain many PE2,3,4 entries (e.g., 85 non-GPCR transmembrane, 69 Zn finger and 33 keratin-associated proteins remain PE2,3,4 s (Fig. 3a) . When Chr distribution of the most resistant (ORs) or most discovered (Zn-finger proteins) groups were plotted (Fig. 3b) , difficult-to-find ORs mapped mostly to Chr 11 (~55%) and Zn-finger proteins mapped mostly to Chr 19 (~55%). Our data demonstrate that 46% of all Chr 19 proteins elevated to PE1 were Zn-finger family members, illustrating this protein family has been highly amenable to confident MS detection over the decade. Therefore, it was not surprising that most Zn-finger protein members were coded on Chr 19 (i.e., 255/698 total Zn-finger genes), which has the highest Chr PE1 reclassification rate over the decade (Fig. 3b) . These data suggest that gene family duplication on particular Chrs explains why some families are resistant or sensitive to PE1 reclassification. In agreement, PE1 discovery has occurred productively, but not uniformly, across all chromosomes (Fig. 4) . Missing protein decadal upgrade statistics range from 16% for Chr Y up to 29% for Chr 1, with raw data representing 425 protein entries for Chr 1 down to only 5 for Chr Y. A higher percentage of Chr 19 proteins (29%) ascended to PE1 compared to those on Chrs 11, 14, 21 or Y (<17%). The HPA chromosome viewer (www.proteinatlas.org/humanproteome/proteinevidence) illustrates many recently-evolved protein family members are present on Chr Y, many ORs on Chr 11 and many keratinassociated proteins on Chr 21. Proteins on these three Chrs have proven relatively resistant to PE1 reassignment. Notably, there was just a single Mt missing protein in 2011 and over the decade all Mt protein-coding genes have now been identified. PE1 upgrade success by the 25 C-HPP teams could be predetermined by the presence of resistant or more easily-identified missing protein families on particular Chrs. However, the current HPP KB processing pipeline utilizing PeptideAtlas, MassIVE and neXtProt (Fig. 1b) makes it impossible to isolate decadal PE1 contributions from any particular C-HPP team as opposed to the overall community. Although the HPP explores capturing full data provenance (i.e., from quantification/identification back to original data source) for FAIR data practices 17 , we can only historically estimate PE1s emanating from community deposition. In silico analysis reveals only 22 human proteins cannot produce the characteristic >2 high-stringency proteotypic peptides of the required length after tryptic digestion 36 . However, many missing proteins may be present at levels below detection limits or in under-studied cell types/tissues, expressed under particular conditions (e.g., stress/infection) or only found in developmental stages (e.g., embryo/fetus) 77 . Equally, difficult-tosolubilize, hydrophobic multi-transmembrane domain membrane proteins may only generate short tryptic peptides that do not meet high-stringency guidelines or are indistinguishable from other family member sequences 76 . Furthermore, transmembrane protein regions cut at single sites are unlikely to release embedded hydrophobic membrane-anchored protein strands 76 . We anticipate that finding the 9.6% remaining PE2,3,4 missing proteins will require exceptional future effort, including careful sampling of rare cells/tissues 78 combined with better sample fractionation and improved detection limits. Low abundance proteins might be enriched using Abs prior to MS. To this end, the HPA has developed Abs against proteotypic sequences in many missing proteins 79 . Several other labs are working on improved protocols for insoluble keratin-associated cross-linked missing proteins, non-tryptic or chemical digestion strategies to increase proteotypic peptide productivity 78,80 , higher efficiency search engines 81 , and compendia of missing protein biological evidence (e.g., MissingProteinPedia) 71 . Future HPP projects anticipate a shift to detection of biologically functional proteoforms 82 , noting their numbers are far larger and more difficult to measure 83 because of heterogeneous nuclear RNA splicing, many PTMs and detection of peptides with single amino acid variants (SAAVs). Considerable PTM and splicing isoform data are already available through neXtProt, including 190,938 PTM sites and 9,193,365 SAAVs 84 . Community impact on the human proteome One barometer of community engagement in the HPP is the magnitude of investigator-submitted raw MS data that have been re-analysed. Many journals require raw data submission and HPP action was a significant factor in journals adopting requirements aligned to HPP data deposition guidelines. Raw data deposition occurs through PX, which registers and standardizes capture/dissemination of public MS data from partner repositories, including founders PRIDE 62 and PeptideAtlas and recent members Mas-sIVE 64 , jPOST 85 , iProX 86 and Panorama Public 87 . As of 2020, a total of 4634 human MS datasets have been received. Each PX dataset is branded with a unique PXD identifier with depositors, publications and voluntary metadata noted 88, 89 . Illustrating the magnitude of this community data,~470 TBs of data have come from 5658 human datasets (~47% of 1 petabyte PRIDE volume), with only 358 (6.3%) of these specifically tagged by depositors as from the HPP. The HPP encourages raw human MS data/metadata submission (including association to HPP) through PX, and that journals request that PXD identifiers be published in accordance with FAIR principles 17 as discussed above. To provide an additional measure of global scientific impact, HUPO commissioned the website construction of the Human Proteome Reference Library (HPRL; https://hupo.org/HPP-HPRL/), where all HPP-associated PubMed searches are hyperlinked and can be accessed and re-run routinely by the community. These hyperlinked searches automatically produce the latest PubMed outputs in a manner where all PubMed filtering, ranking and timeline tools can be applied subsequently by The HPRL can also be used to measure impact (Fig. 5) , where PubMed identifiers (PMIDs) from 2010 to 2019 references can be captured using web APIs or software, e.g., NCBI E-utilities like https://www.ncbi.nlm.nih.gov/books/NBK25501 or the easy-PubMed R interface (www.rdrr.io/cran/easyPubMed) that enable extraction and aggregation of PubMed bibliometric records. Paralleling the low percentage of PX datasets tagged as emanating from the HPP, HPRL data show only~1000 PubMed title and abstract searches that specifically and cumulatively mention either the terms HPP, C-HPP and/or B/D-HPP (or their unabbreviated counterparts) out of a total of more than 50,000 'human AND proteomics' title/abstract search outcomes over the decade. In contrast to this and more reassuringly, HPRL searches of the community's biology/ or disease studies covered by nine of the B/D-HPP teams generated >5000 publications each since 2010 (Fig. 5b) , with 'cancer proteomics' topping rankings at around 20,000 publications. Although this may seem surprising, a PubMed search of 'human genome project' produced a similarly modest number of~1900 hits during the HGP's first decade. To visualize outputs from various HPP teams and the proteomics community in general, we have employed VOSviewer 91 to construct and visualize bibliometric networks from collective PubMed searches (example shown in Fig. 5c ). This particular analysis revealed a remarkable number of highly-interconnected relationships that have evolved between HPP teams, pillars and initiatives-providing evidence for the impressive level of established international collaboration developed over the decade. A key aspect of biomedical research lies in translating discovery into clinical use. Protein assays remain a cornerstone of diagnostics. Although individual proteins can be measured diagnostically with high precision (i.e., sensitivity and specificity), some assays suffer low specificity due to cross-reactivity with interfering substances including autoantibodies (e.g., thyroglobulin immunoassays). Modern SRM/MRM/PRM assays allow multiple proteins to be measured simultaneously, accurately, sensitively and with high specificity. In addition, the use of liquid chromatography MS with immunocapture assays has been reported to eliminate interferences 92 . Moreover, as most diseases are heterogeneous and multigenic, it is likely that multiplexed proteomic or multi-omics panels will achieve higher accuracy (e.g., optimized biomarkers for ovarian malignancy with adnexal masses 93 ). The HPP assists in the development of proteomic educational programmes with pathology societies to train the pathology community on the potential impact of proteomic technologies. Below, in recognition of the impact human proteomics can and is having in precision medicine, we highlight examples demonstrating the role of the HPP and proteomics in tackling contemporary medical grand challenges. Cancer precision medicine. As mentioned above, PubMed extracts~20,000 published human cancer proteome studies since 2011 (Fig. 5b) . Although genomics can routinely determine highrisk, predisposition and aspects related to tumour burden and recurrence, effective targeted cancer treatment is still not available for all cancers. For example, systematic genome-wide studies like the Pan-Cancer Analysis integrated analyses of >2600 whole genomes from 38 tumour types with matching normal tissues, uncovered many cancer-associated genes 94 , chromosome rearrangements, some unknown drivers but few new therapeutic targets. This is mainly because mutations do not automatically cause predicted changes in the proteome, making it difficult to establish which changes are crucial biochemical drivers from those that are not. Integrating genomic and proteomic data (i.e., proteogenomics) has the potential to provide insights into causes and mechanisms underlying diseases, including the hallmarks of cancer biology 95 . This can facilitate the implementation of effective therapeutic intervention. The value of a proteogenomic analysis of functional consequences of cancer somatic mutations has assisted in narrowing down candidate driver genes within large deletions and amplified regions 96 . Reviewing the underlying causes of breast cancer also demonstrates that coupling genomic/transcriptomic data with proteomic/phosphoproteomic analysis was more insightful than any individual approach. Of note, melanoma tumour genomic BRAF driver mutations match corresponding protein sequences 97 , illustrating that proteomic landscapes add value to genomic data, when considered with patient tumour histopathology and clinical metadata 98 . Proteomics substantially benefits a comprehensive understanding of precision medicine. To illustrate this, NCI's Clinical Proteomic Tumor Analysis Consortium (CPTAC; proteomics. cancer.gov) in collaboration with the B/D-HPP cancer team, established guidelines, data sharing, and standards in analytical and computational workflows to ensure rigour in designing and performing research [99] [100] [101] [102] [103] . CPTAC applied these standards and workflows to tumours previously genomically characterized by The Cancer Genome Atlas. In doing so, CPTAC pioneered the integration of proteomics with genomics (i.e., proteogenomics) to produce a more unified and comprehensive understanding of cancer biology and implemented its transition into cancer clinical research studies 96, 104, 105 . Largely due to these efforts, NCI hosts open-access repositories of unified proteogenomics datasets, assays and reagents, including the Proteomic Data Commons (pdc.cancer.gov), a fit-for-purpose targeted assay site (assays. cancer.gov) 102, 106 and an Ab portal (antibodies.cancer.gov) 107 . These activities empowered the recent creation of the International Proteogenome Consortium (ICPC; icpc.cancer.gov) 108 . Collectively, CPTAC and ICPC collaborators have comprehensively characterized 13 cancer types at the proteogenomics level, with all datasets publicly accessible [109] [110] [111] [112] [113] [114] [115] [116] [117] . Cardiovascular diseases. Cardiovascular disease (CVD) research is challenged by daunting structural heterogeneity and molecular complexity. For example, cardiac circuitry function/dysfunction cannot be reduced to differentially expressed single genes. On the other hand, proteogenomics allows assessment of interactions, pathways and networks and informs diagnosis and therapy of multifactorial CVDs. Over the decade, CVD proteomics has broadened from identifying single canonical proteins to mapping proteoforms derived from combinations of alternative splicing, cleavage, and PTMs 33, 83 . Now, identification of proteoforms (e.g., genetic variations, alternatively spliced products, phosphorylation 118 , glycosylation 119 , oxidative 120 and other PTMs 121 ) allow CVD sub-classification. In parallel, the heart is uniquely sensitive to alternative splicing (frequently altered in congenital heart diseases), explaining the B/D-HPP's interest in developing assays for splice isoform-specific changes in cardiomyocyte development and maturation 122, 123 . Many technological developments have been prominent, including phospho-PTM analysis to identify PDE5A targets in heart failure therapy 118 and proximity labelling to assess protein-protein interactions involved in β-adrenergic signalling of contractility in cardiac fight-or-flight responses 124 and evaluation of regenerative stem cell therapy efficacy in postinfarct hearts 125 . Proteomic studies have also addressed the kilometres of vascular beds and extracellular matrix responsible for transporting blood, giving valuable insights into the molecular anatomy of aneurysms and atherosclerosis 126, 127 . Targeted MS methods have been developed, again especially where interferences obfuscate immunoassays 128 or where additional biological context is required 129 . Likewise, volumetric absorptive microsampling VAMS blood collection devices (e.g., Mitra devices or dried blood spots) allow patients to mail samples from home to analytical labs to undertake SRM/MRM/PRM assays quantifying CVD risk-associated apolipoproteins 130 or other markers 131 . In summary, consumer-based CVD proteomics-based precision medicine testing services are now coming of age. Microorganism detection. Proteomics and the HPP have made fundamental contributions to understanding pathogenic infection, providing diagnostics and developing therapies 132 . The B/D-HPP Infectious Diseases team promotes international proteomics collaborations investigating viral, bacterial, fungal and parasitic diseases. MALDI-time-of-flight (TOF)-MS, once considered revolutionary, is now established as a routine tool in clinical microbiology 133 . Classical phenotypic tests identify unknown and potentially pathogenic microorganisms, but may require incubation for several days, with misidentification resulting in adverse treatment consequences. MALDI-TOF-MS provides significantly shortened analyses (now minutes) with improved accuracy on single colony or bacterial pellets for difficult-todetect microorganisms, using automated spectra acquisition and extensive reference spectra databases 134 . Minor spectral differences enable typing below species levels 135 , allowing subspecies identification through epidemiological analyses. Bacteria and yeasts (most clinical identifications), mycobacteria 136 and moulds 137 can now be identified accurately and rapidly. Further MS clinical diagnostic applications are being investigated (e.g., antibiotic resistance (ART) and susceptibility testing (AST) based on hydrolytic β-lactamase activity 138 ), with kits under development commercially through STAR-Carba, STAR-Cepha and Bruker Daltonics. SARS-CoV-2 virology. The recent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) outbreak that causes COVID-19 disease represents a major threat to human health and our economies [139] [140] [141] . The pandemic underscores our need to understand virus pathobiology, identify host-pathogen interactions that support replication, find biomarkers correlative with clinical outcome and expand surveillance. Many omics studies followed the 2003 SARS-CoV-1 and related MERS and IBV coronavirus outbreaks [142] [143] [144] [145] [146] . The cell surface receptor for the CoV-1 and CoV-2 surface spike protein has been identified by affinity-MS to be angiotensin-converting enzyme 2 (ACE2) 147 , which in a recent large-scale study based on antibodybased proteomics was shown to be mainly localized to the digestive system, kidney, heart, testis, placenta, eye and upper respiratory epithelia 148 . Virus binding leads to proteolysis by the transmembrane serine protease TMPRSS2 expressed in airway epithelia 149 , thus a clinically approved TMPRSS2 inhibitor (camostat mesylate) is being investigated to block infections 150, 151 . Furthermore, proteomics has characterized the infectious CoV-1 viral particle 143 , temporal changes in host cells during infection 142 and virusinduced endoplasmic reticulum membrane remodelling into double-membrane vesicles 152,153 that house viral replication compartments 146 . Proximity labelling revealed >500 host and 14 viral protein associations with the viral replicase NPS2 146 , highlighting vesicular trafficking, autophagy and splicing proteins in coronavirus replication, which if shown to also be the case for CoV-2, indicate potential drug targets. Building on this knowledge of coronavirus infection, recent proteomics studies have focused on SARS-CoV-2, uncovering additional potential therapeutic targets 154, 155 . MS and array-based proteomics serology has screened for potential biomarkers and Abs against infection 156, 157 . Clinical isolate infection models have been developed using Caco-2 cells 154 with temporal proteome changes identified during infection using multiplexed MS by combining metabolic labelling with tandem mass tagging methods. Consistently, host vesicular trafficking, translation, RNA splicing, nucleotide synthesis and glycolysis pathway proteins were upregulated following infection 143, 144 and targeting these processes with inhibitors revealed potential therapeutic targets 158, 159 . Additionally, affinity-MS interactome studies examined 26 of 29 total SARS-CoV-2 proteins expressed within HEK293T human cells 155 , suggesting 69 existing drugs merit further investigation. Moreover, a recent phosphoproteome analysis pointed to the regulation of viral proteins through PTMs 160 . The SARS-CoV-2 pandemic highlights the need for applying proteomic approaches to the development of serologic testing and preclinical and computational model systems to evaluate patient responses to infection 156 . Serological biomarkers of asymptomatic/ symptomatic infection, disease severity, risk of re-infection and/or vaccine efficacy are being characterized 157, 161 . In addition to this accumulating omics knowledge, many aspects of SARS-CoV-2 pathobiology await further exploration including development of additional methods for clinical virus detection, identification of infection stage, and an in-depth understanding of functional spatiotemporal virus-host protein interactions and organelle remodelling [162] [163] [164] . For example, recent studies that utilize targeted MS for SARS-CoV-2 protein detection and proteomic characterization of serological immune responses 161,165-167 from patient samples may bolster PCR screening for the assessment of disease severity 168 . Other proteomics approaches can also be deployed to further expand the understanding of SARS-CoV-2 biology. Among these is the application of TAILS N-terminomics that promises to identify many SARS-CoV-2 protease substrates and those cellular pathways inactivated by viral proteolysis, as reported for other viruses 169 . In summary, proteomics plays increasingly important roles in understanding viral outbreak biology, accurate diagnosis and effective treatment and is positioned to continue to co-ordinate and drive international collaborations towards these goals. Western and Eastern cultures urge us to know thyself and thy enemy. These axioms resonate with precision medicine where future benefits arise from a detailed omics understanding of the hallmarks of health and disease. Here, we reviewed the construction of a community-endorsed, high-stringency blueprint of the human proteome. The decadal neXtProt HPP PE metrics shown here demonstrate the community's progressive success in PE1 identification from 13,588 in 2011 to 17,874 PE1s in 2020, marking the completion of >90% of the human proteome parts list (see strategic objective 1 above). We also present specific examples demonstrating proteomics will be an integrated component (with genomics and other omics) in future biomedical science discovery and precision medicine. HUPO recommits to its original HPP strategic aims as well as the FAIR data principles 17 , while anticipating the following future priorities: 1. Unearth credible proteomics data for the majority of current PE2 proteins: Since most PE1 identifications come from former PE2s, our future strategy is to find credible data for 95-99% of current PE2s, allowing reclassification of these to PE1. PE3 also remain promising, as homologous proteins are detected in related species. 2. Unravel currently unknown proteome functionalities: Fill functional annotation gaps for all protein-coding elements, with a priority on credibly identified proteins 170 and develop, expand and apply function prediction tools 24, 171, 172 . 3. Expand the HPP KB: Maintain a sustainable knowledgetransfer HPP KB infrastructure with funding that captures/ displays high-stringency partner omics data streams and publication data (HPRL) to researchers and the public in an accessible and compelling manner. 4. Develop community-approved multi-omics technology guidelines: Explore DIA-MS, Ab and aptamer-based multiplexed assays, top-down MS and other not yet invented multi-omics technologies. 5. Champion collaborative multi-omics health/disease approaches: In addition to extensions in HPP KB partnerships, the HPP will collaborate with international and regional initiatives (including Human Variome, Human Cell Atlas, hPOP/iPOP, MoTrPAC, HuBMAP, Cancer Moonshot, HTAN, EDRN, CPTAC, ICPC and upcoming international initiatives) around multi-omics approaches to human disease processes, biomarker discovery and therapeutic development. 6. Apply and improve single-cell proteomic technology: Develop technologies that allow detection and quantification of proteomes in single cells to further understand cellular/tissue heterogeneity, differentiation, diseases and the intrinsic biological noise in health and disease 173, 174 . Many single-cell proteomics advances will be explored to analyse cellular heterogeneity [175] [176] [177] [178] . Sensitivity increases (analogous to PCR) and trade-offs maximizing coverage per cell with throughput/accuracy will be studied 179 . 7. Champion dual Ab-capture with MS identification: Enhance the accuracy of Ab-based epitope/antigen detection by confirmation using high-accuracy, high-stringency MS identifications across real-life spatio-temporal biological settings. Collectively, there is recognition within the HPP that future stringent co-registration of MS with Ab data are required to achieve the ultimate spatio-temporal human proteome expression atlas 180 . 8. Exploit massively parallel MS 174 to increase throughput: For rapid, sensitive and higher-throughput MS analysis. 9. Capitalize on human disease biobanks: The HPP will work with biobanking consortia to improve access for proteomics researchers to highly curated and accurately annotated clinical samples collected in a standardized manner. 10. Encourage higher levels of community engagement: The HPP will continue to reach out to the community, supporting findable, accessible, interoperable and reusable (FAIR) data principles 17 , while encouraging and appropriately recognizing all contributions to the re-analysis of proteomics data. The post SARS-CoV-2 pandemic world will be different. It is likely that new paradigms to accelerate precision medicine will emerge. These will undoubtedly involve global collaboration (even between competing entities) using multidisciplinary approaches that enable the fast-tracking of novel diagnostic tests and precision therapeutics. Almost certainly these outcomes will require knowledge involving the human proteome-celebrated here in the inaugural HPP High-Stringency Blueprint. Received: 19 December 2019; Accepted: 25 September 2020; A human proteome project with a beginning and an end Building the 'practical' human proteome project-the next big thing in basic and clinical proteomics Is a gene-centric human proteome project the best way for proteomics to serve biology First HPP publication after launch, outlining systematic global effort to map the human proteome (abundance, distribution, temporal-spatial and subcellular localisation, interactions and function) and decribing reource pillars (MS, Ab and KB), with a biologically-driven and a chromosome-based mapping initiatives to deliver a protein parts list, reagents and tools Human genome. Finally, the book of life and instructions for navigating it Metrics for the human proteome project 2013-2014 and strategies for finding missing proteins Standard guidelines for the chromosome-centric human proteome project A first step toward completion of a genome-wide characterization of the human proteome 2019 annual HPP metrics publication (NeXtProt 2019-01-11) reporting progress in credibly identifying and characterising the human proteome Details current high-stringency MS guidelines criteria used by the HPP for PE1 status (evidence for protein existence), where only protein entries with two or more neXtProt uniquely-mapping, non-nested peptides with length 9 amino acids or greater are deemed to have sufficient evidence Trans-Proteomic Pipeline, a standardized data processing pipeline for large-scale reproducible proteomics informatics The neXtProt peptide uniqueness checker: a tool for the proteomics community Looking for missing proteins in the proteome of human spermatozoa: an update On the proper use of mass accuracy in proteomics Precision proteomics: the case for high resolution and high mass accuracy Human proteome project mass spectrometry data interpretation guidelines 2.1 The FAIR guiding principles for scientific data management and stewardship A draft map of the human proteome Mass-spectrometry-based draft of the human proteome Analyzing the first drafts of the human proteome Flexible and fast mapping of peptides to a proteome with ProteoMapper The chromosome-centric human proteome project for cataloging proteins encoded in the genome Advances in the chromosome-centric human proteome project: looking to the future Launching the C-HPP neXt-CP50 pilot project for functional characterization of identified proteins with no known function Human liver proteome project: plan, progress, and perspectives A cell-type-resolved liver proteome Proteomics identifies new therapeutic targets of early-stage hepatocellular carcinoma Integrated multiomics example (proteome, phosphoproteome, transcriptome, and whole-exome sequencing) analyses leading to clinical outcomes, performed on 103 lung adenocarcinomas, revealing many cancer-associated features (e.g., tumourassociated variants, patient clinical outcome correlation with EGFR or TP53 mutational status, proteomic stratification of lung cancer subtypes and potential drug targets Proteome-wide profiling of activated transcription factors with a concatenated tandem array of transcription factor response elements A region-resolved mucosa proteome of the human stomach Antigen presentation profiling reveals recognition of lymphoma immunoglobulin neoantigens The immunopeptidomic landscape of ovarian carcinomas Precision profiling of the cardiovascular post-translationally modified proteome: where there is a will, there is a way In vivo phosphoproteome analysis reveals kinome reprogramming in hepatocellular carcinoma Protein s-nitrosylation controls glycogen synthase kinase 3β function independent of its phosphorylation state Generation and verification of a compendium of highly specific SRM assays that enable quantification of >95% of all annotated human proteins. Provides definitive MS data on 166,174 proteotypic peptides that facilitate the design of multiple, independent assays to quantify most human proteins, numerous splice-variants, non-synonymous mutations Identification and validation of human missing proteins and peptides in public proteome databases: data mining strategy Multi-laboratory assessment of reproducibility, qualitative and quantitative performance of SWATH-mass spectrometry A human protein atlas for normal and cancer tissues based on antibody proteomics Spatial proteomics: a powerful discovery tool for cell biology Getting to know the neighborhood: using proximity-dependent biotinylation to characterize protein complexes and map organelles Antibodypedia, a portal for sharing antibody and antigen validation data Towards a knowledge-based human protein atlas A genome-wide transcriptomic analysis of protein-coding genes in human blood cells An atlas of the protein-coding genes in the human, pig, and mouse brain A subcellular map of the human proteome A pathology atlas of the human cancer transcriptome An atlas of human metabolism First of many encyclopedic HPA atlases that map the spatio-temporal expression human proteome based on integrated multi-omics data involving quantitative transcriptomics and Ab microarraybased tissue immunohistochemistry and delivering spatial localisation of human proteins down to the single-cell level A proposal for validation of antibodies Enhanced validation of antibodies for research applications Deep learning is combined with massive-scale citizen science to improve large-scale image classification Analysis of the Human Protein Atlas Image Classification competition The ELIXIR Core Data Resources: fundamental infrastructure for the life sciences Integration of transcriptomics and antibody-based proteomics for exploration of proteins expressed in specialized tissues Cell type-specific expression of testis elevated genes based on transcriptomics and antibody-based proteomics The human cell atlas The minimum information about a proteomics experiment (MIAPE) neXtProt: a knowledge platform for human proteins The neXtProt knowledgebase in 2020: data, tools and usability improvements neXtProt: organizing protein knowledge in the context of human proteome projects The PRIDE database and related tools and resources in 2019: improving support for quantification data State of the human proteome in 2014/2015 as viewed through PeptideAtlas: enhancing accuracy and coverage through the AtlasProphet ProteinExplorer: a repository-scale resource for exploration of protein detection in public mass spectrometry data sets The neXtProt knowledgebase on human proteins: 2017 update Comet: an open-source MS/MS sequence database search tool X!!Tandem, an improved method for running X!tandem in parallel on collections of commodity computers Development and validation of a spectral library searching method for peptide identification from MS/MS MS-GF+ makes progress towards a universal database search tool for proteomics Assembling the community-scale discoverable human proteome Community encouragement to identify biological data that complement high-stringency MS strategies to accelerate discovery and understanding of human proteome PE2,3,4 missing proteins. Database allows unpublished, preliminary or proprietary data (e.g., antibody, MS, cell biology and genetic studies CHESS: a new human gene catalog curated from thousands of large-scale RNA sequencing experiments reveals extensive transcriptional noise Dynamic visualization of multi-level molecular data: the Director package in R Can we predict protein from mRNA levels? Proteomic and interactomic insights into the molecular basis of cell functional diversity In silico peptide repertoire of human olfactory receptor proteomes on high-stringency mass spectrometry Toward completion of the human proteome parts list: progress uncovering proteins that are missing or have unknown function and developing analytical methods Proteomic and N-terminomic TAILS analyses of human alveolar bone proteins: improved protein extraction methodology and LysargiNase digestion strategies increase proteome coverage and missing protein identification The human protein atlas: a spatial map of the human proteome Multiproteases combined with high-pH reverse-phase separation strategy verified fourteen missing proteins in human testis tissue Open-pFind enhances the identification of missing proteins from human testis tissue Proteoform: a single term describing protein complexity How many human proteoforms are there? What will neXtProt help us achieve in 2020 and beyond The jPOST environment: an integrated proteomics data repository and database iProX: an integrated proteome resource Panorama public: a public repository for quantitative data sets processed in Skyline ProteomeXchange provides globally coordinated proteomics data submission and dissemination The ProteomeXchange consortium in 2020: enabling 'big data' approaches in proteomics Structural genomics: keystone for a Human Proteome Project Software survey: VOSviewer, a computer program for bibliometric mapping Measurement of thyroglobulin by liquid chromatography-tandem mass spectrometry in serum and plasma in the presence of antithyroglobulin autoantibodies The road from discovery to clinical diagnostics: lessons learned from the first FDA-cleared in vitro diagnostic multivariate index assay of proteomic biomarkers Pan-cancer analysis of whole genomes Details the hallmarks of cancer acquired during multistep development of human tumors (i.e., sustaining proliferative signaling, evading growth suppressors, resisting cell death, enabling replicative immortality, inducing angiogenesis, activating invasion/ metastasis, genome instability, inflammation, reprogramming of energy metabolism and evading immune destruction). Creates a roadmap for the understanding of multi-omics based cancer data Proteogenomics connects somatic mutations to signalling in breast cancer The hidden story of heterogeneous B-raf V600E mutation quantitative protein expression in metastatic melanoma-association with clinical outcome and tumor phenotypes Clinical protein science in translational medicine targeting malignant melanoma Repeatability and reproducibility in proteomic identifications by liquid chromatography-tandem mass spectrometry Multi-site assessment of the precision and reproducibility of multiple reaction monitoring-based measurements of proteins in plasma Demonstrating the feasibility of large-scale development of standardized assays to quantify human proteins Targeted peptide measurements in biology and medicine: best practices for mass spectrometry-based assay development using a fit-forpurpose approach Protein-based multiplex assays: mock presubmissions to the US Food and Drug Administration Proteogenomic characterization of human colon and rectal cancer Integrated proteogenomic characterization of human highgrade serous ovarian cancer CPTAC assay portal: a repository of targeted proteomic assays Antibody characterisation-an essential researchers' resource Revolutionizing precision oncology through collaborative proteogenomics and data sharing Integrated proteogenomic characterization of clear cell renal cell carcinoma Proteogenomic characterization of endometrial carcinoma Proteogenomic characterization reveals therapeutic vulnerabilities in lung adenocarcinoma Example of a CPTAC/ICPC proteogenomics study highlighting new therapeutic insights into colorectal cancer. Comparative proteomic and phosphoproteomic analysis of a prospectively-collected paired colon cancer tumour and adjacent normal tissue catalogue of colon cancer-associated proteins and phosphorylation sites, including known and putative new biomarkers Proteogenomic characterization of ovarian HGSC implicates mitotic kinases, replication stress in observed chromosomal instability APOBEC3A is an oral cancer prognostic biomarker in Taiwanese carriers of an APOBEC deletion polymorphism Proteogenomic characterization of human early-onset gastric cancer Integrated proteogenomic characterization of HBV-related hepatocellular carcinoma Proteogenomics of non-smoking lung cancer in East Asia delineates molecular signatures of pathogenesis and progression Demonstrates the potential of proteomics and PTM analyses to increase our understanding of hypertrophic heart disease. Transcription factor activation and phosphoproteome analyses of myocytes revealed PDE9A regulates cGMP signalling independent of the NO-pathway in stress-induced heart disease Reference glycan structure libraries of primary human cardiomyocytes and pluripotent stem cell-derived cardiomyocytes reveal cell-type and culture stage-specific glycan phenotypes Integrated dissection of cysteine oxidative post-translational modification proteome during cardiac hypertrophy Cardiac troponins may be irreversibly modified by glycation: novel potential mechanisms of cardiac performance modulation Splice-junction-based mapping of alternative isoforms in the human proteome An unbiased proteomics method to assess the maturation of human pluripotent stem cell-derived cardiomyocytes Mechanism of adrenergic CaV1.2 stimulation revealed by proximity proteomics Cardiopoietic stem cell therapy restores infarction-altered cardiac proteome Glycoproteomic analysis of the aortic extracellular matrix in Marfan patients Extracellular matrix proteomics identifies molecular signature of symptomatic carotid plaques Targeted aproach to distinguish and determine absolute levels of GDF8 and GDF11 in mouse serum Protein markers and risk of type 2 diabetes and prediabetes: a targeted proteomics approach in the KORA F4/FF4 study Application of volumetric absorptive microsampling for robust, high-throughput mass spectrometric quantification of circulating protein biomarkers Concentration determination of >200 proteins in dried blood spots for biomarker discovery and validation The impact of mass spectrometrybased proteomics on fundamental discoveries in virology Ongoing revolution in bacteriology: routine identification of bacteria by matrix-assisted laser desorption ionization time-of-flight mass spectrometry How MALDI-TOF mass spectrometry can aid the diagnosis of hard-to-identify pathogenic bacteria-the rare and the unknown Can MALDI-TOF mass spectrometry reasonably type bacteria? How to identify non-tuberculous Mycobacterium species using MALDI-TOF mass spectrometry Identification of molds by matrix-assisted laser desorption ionization-time of flight mass spectrometry Susceptibility testing of bacteria using MALDI-TOF mass spectrometry A new coronavirus associated with human respiratory disease in China An interactive web-based dashboard to track COVID-19 in real time How will country-based mitigation measures influence the course of the COVID-19 epidemic? Quantitative proteomics analysis reveals BAG3 as a potential target to suppress severe acute respiratory syndrome coronavirus replication Proteomics analysis unravels the functional repertoire of coronavirus nonstructural protein 3 The cellular interactome of the coronavirus infectious bronchitis virus nucleocapsid protein and functional implications for virus biology MERS-CoV and H5N1 influenza virus antagonize antigen presentation by altering the epigenetic landscape Determination of host proteins composing the microenvironment of coronavirus replicase complexes by proximity-labeling Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus The protein expression profile of ACE2 in human tissues Efficient activation of the severe acute respiratory syndrome coronavirus spike protein by the transmembrane protease TMPRSS2 The spike glycoprotein of the new coronavirus 2019-nCoV contains a furin-like cleavage site absent in CoV of the same clade Illustrates the rapidity at which modern proteomics can contribute to the understanding of a recent pandemic, revealing insights into viral transmission and therapeutic targets. Indicates SARS-CoV-2 uses the receptor ACE2 for human cell entry and serine protease TMPRSS2 for SARS-CoV-2 spike protein priming. A clinically approved TMPRSS2 inhibitor blocked virus entry and sera from convalescent SARS-CoV-2 patients crossneutralized viral entry Making of viral replication organelles by remodeling interior membranes Proteomic approaches to uncovering virus-host protein interactions during the progression of viral infection Proteomics of SARS-CoV-2-infected host cells reveals therapy targets A SARS-CoV-2 protein interaction map reveals targets for drug repurposing SARS-CoV-2 proteome microarray for global profiling of COVID-19 specific IgG and IgM responses Proteomic and metabolomic characterization of COVID-19 patient sera Inhibitory effect of mizoribine and ribavirin on the replication of severe acute respiratory syndrome (SARS)-associated coronavirus Nucleoside analogues for the treatment of coronavirus infections Characterisation of the transcriptome and proteome of SARS-CoV-2 reveals a cell passage induced in-frame deletion of the furin-like cleavage site from the spike glycoprotein Ultra-high-throughput clinical proteomics reveals classifiers of COVID-19 infection Post-translational modification control of innate immunity Orchestration of protein acetylation as a toggle for cellular defense and virus replication Temporal dynamics of protein complex formation and dissociation during human cytomegalovirus infection Mass spectrometric identification of SARS-CoV-2 proteins from gargle solution samples of COVID-19 patients Proteotyping SARS-CoV-2 virus from nasopharyngeal swabs: a proof-of-concept focused on a 3 min mass spectrometry window A portrait of the human organelle proteome in space and time during cytomegalovirus infection Multiplex targeted mass spectrometry assay for one-shot flavivirus diagnosis N-Terminomics TAILS identifies host cell substrates of poliovirus and coxsackievirus B3 3C proteinases that modulate virus infection FusionPro, a versatile proteogenomic tool for identification of novel fusion transcripts and their potential translation products in cancer cells Structure and protein interaction-based gene ontology annotations reveal likely functions of uncharacterized proteins on human chromosome 17 Blinded testing of function annotation for uPE1 proteins by I-TASSER/COFACTOR pipeline using the 2018-2019 additions to neXtProt and the CAFA3 challenge Single-cell analysis tools for drug discovery and development National Cancer Institute think-tank meeting report on proteomic cartography and biomarkers at the single-cell Level:interrogation of premalignant lesions Single-Cell Omics Single-cell mass-spectrometry quantifies the emergence of macrophage heterogeneity Multiplexed single cell protein expression analysis in solid tumours using a miniaturised microfluidic assay Single-cell proteomics reveal that quantitative changes in coexpressed lineage-specific transcription factors determine cell fate Transformative opportunities for single-cell proteomics Investigating the applicability of antibodies generated within the human protein atlas as capture agents in immunoenrichment coupled to mass spectrometry was corresponding author, conceptualized HPP decadal blueprint, HUPO proteomics timeline and HPRL, and he designed/integrated all aspects of the manuscript made academic co-contributions to manuscript section drafting, editing, review and/or proof-reading Health and Human Sciences Australia. 3 Institute for Systems Biology Seoul 120-749, South Korea. 8 Faculty of Dentistry USA. 12 Science for Life Laboratory, School of Engineering Sciences in Chemistry, Biotechnology and Health, KTH Royal Institute of Technology, 17121 Solna, Sweden. 13 Rudbeck Laboratory, Department of Immunology 23 Cancer Biomarkers Research Branch, National Cancer Institute, National Institutes of Health, 9609 Medical Center Drive, Suite 5E136 Wellcome Trust Genome Campus, Hinxton, Cambridge CB10 1SD, UK. 30 School of Biotechnology and Biomolecular Sciences HUPO acknowledges collaborators, proteomic scientists, independent partners, industry vendors and members of the scientific community who have contributed to the HPP. A full alphabetical listing of the Human Proteome Project members appears in the Supplementary Information. In recognition of the many accomplishments, HUPO has produced a publicly available HPP timeline available through https://hupo.org/ Proteomics-Timeline to be released with this HPP Blueprint. Parts of this work were supported by grants to S.R.P. is founder and chief scientific officer of Atturos, a clinical diagnostics company. M.K. is an employee of Bruker Daltonics, a manufacturer of MS systems. All other authors declare no competing interests. Supplementary information is available for this paper at https://doi.org/10.1038/s41467-020-19045-9.Correspondence and requests for materials should be addressed to M.S.B. Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/.