6 min read
How many proteins are there?
Humans have 19,442 protein-coding genes, 172,117 annotated translated products, and billions to trillions of individual protein molecules per cell. Across all species, 475 million unique sequence clusters have been catalogued.

Matic Broz Computational chemist

There is no single protein count because “how many proteins” can refer to protein-coding genes, distinct translated products, protein types, or individual molecules.
For humans, the clearest reference count is 19,442 protein-coding genes. Current annotations contain 172,117 distinct translated products, while each cell contains billions to trillions of individual protein molecules. Across all species, researchers have catalogued about 475 million unique protein sequence clusters.
How many proteins does a human have?
Humans have 19,442 protein-coding genes, but that does not mean the human body contains only 19,442 proteins.[2]
These human genes can produce 278,455 protein-coding transcripts, which encode 172,117 distinct translated protein products. The count changes depending on whether the question refers to genes, transcripts, translated products, or experimentally detected proteins.[2]
The Human Proteome Project (HPP) asks a different question: how much of the reference proteome has been directly observed in experiments? The 2025 HUPO HPP report defines a reference proteome of 19,435 proteins and reports that 94% of them have been confidently detected with protein-level evidence.[3][4]
| Level | Count | Meaning |
|---|---|---|
| Protein-coding genes | 19,442 | GENCODE v50 coding genes |
| Coding transcripts | 278,455 | Protein-coding transcript isoforms |
| Translations | 172,117 | Distinct translated products |
| HPP reference | 19,435 proteins | 2025 HUPO HPP target list |
| Detected (PE1) | 94% | HPP share with confident evidence |
A gene count is not a protein count. One gene can produce multiple protein products through alternative splicing, alternative start sites, and sequence variation. That is why 19,442 genes correspond to 172,117 annotated translations.
The diversity goes further. A proteoform includes different molecular forms of a protein produced by genetic variation, splicing, RNA editing, cleavage, and post-translational modifications such as phosphorylation or glycosylation.[9] The number of possible human proteoforms is far larger than the number of genes or translations, but no fixed count exists because it depends on how variants and modification combinations are defined and measured.
For citation purposes, avoid writing that humans "have 20,000 proteins" unless you mean protein-coding genes or a reference proteome. If you mean molecular forms in cells, use proteoforms and state that the exact count is not fixed.
How many protein sequences have we catalogued?
UniProt's UniRef100 database contains 475,217,233 protein sequence clusters as of June 2026. UniRef groups identical sequences together, so this is the number of unique protein sequences observed and deposited across all species, from bacteria to the human proteome.[1]
UniProt provides two further levels of redundancy reduction, which give different numbers:
| UniRef dataset | What it counts | Count |
|---|---|---|
| UniRef100 | Exact-sequence clusters | 475,217,233 |
| UniRef90 | Clusters at 90% sequence identity | 188,848,220 |
| UniRef50 | Clusters at 50% sequence identity | 60,315,044 |
UniRef90 groups closely related sequences that likely share the same function. UniRef50 compresses the database further into broader protein families. Which number to cite depends on whether you want exact sequences or functional groups.[1]
These numbers describe the protein sequence space that biology has explored so far. The theoretical space is far larger. For a protein just 100 amino acids long, the number of possible sequences is 20100 because each of 100 positions can hold one of 20 standard amino acids. The 475 million catalogued clusters represent only a tiny corner of what chemistry allows.[9]
Experimental protein structures tell a similar story. The Protein Data Bank holds 256,006 experimentally determined 3D structures and an additional 1,062,058 computed structure models, for a total of over 1.3 million available structures as of mid-2026. Most of the computed models come from AlphaFold and represent predicted structures for sequences whose experimental structures have not yet been solved.
How many proteins are in a human cell?
A human cell contains roughly 10 billion to 10 trillion individual protein molecules, although estimates vary with cell type, size, and counting method.[5][7]
An average yeast cell contains about 42 million protein molecules, based on a 2018 study that unified 21 yeast protein-abundance datasets.[8]
A typical mammalian cell - roughly 3,000 µm³ - contains about 10 billion protein molecules, at a density of roughly 3 million molecules per cubic micrometer.[5]
A 2025 NIGMS educational estimate puts the figure at 10 trillion protein molecules for a human cell.[7]
| Measure | Value | Source |
|---|---|---|
| Yeast molecules | ~42 million | 2018 Cell Systems |
| Mammalian molecules | ~10 billion | 2023 Nature Methods |
| Human-cell estimate | ~10 trillion | 2025 NIGMS |
| Protein density | ~3 million per µm³ | 2023 Nature Methods |
| Fast proteomics run | 10,411 | 2024 proteomics study[6] |
These numbers differ by orders of magnitude because they describe different organisms and cell sizes. Yeast cells are much smaller than human cells, and human cells themselves vary enormously - a red blood cell, a neuron, and an oocyte have very different volumes and protein concentrations. Use the organism and cell type whenever possible. "A yeast cell contains about 42 million protein molecules" is precise; "a cell contains 42 million proteins" is not.
These are molecule counts, not counts of unique protein types. A single protein species can exist in thousands to billions of copies inside one cell. Deep mass-spectrometry studies can identify around 10,411 protein groups in a 30-minute human cell-line run, which illustrates the gap between counting protein copies and counting protein species.[6]
What percentage of a cell is protein?
Protein makes up about 18% of the total mass of a typical mammalian cell, or about 60% of its dry mass.[10]
The two percentages use different denominators. A typical mammalian cell is about 70% water and 30% dry material by mass. If protein accounts for 60% of that dry material, it accounts for about 18% of the whole cell: 30% × 60% = 18%.[10]
These are reference values, not constants. The PLOS ONE study lists protein at about 50–55% of dry mass in E. coli and 35–60% in yeast, while mammalian cell composition can shift with cell type and growth conditions.[10] Protein mass also cannot be converted directly into a molecule count because protein molecular weights vary widely.
How to cite this page
Use this article as a dated statistics source only with the access date, because UniRef, GENCODE, UniProtKB, and HPP counts change across releases.
ProteinIQ. "How many proteins are there?" Updated July 22, 2026. Accessed [your access date]. https://proteiniq.io/guides/number-of-proteins
Sources▼
- UniRef REST API counts for UniRef100, UniRef90, and UniRef50 UniProt · June 30, 2026. https://rest.uniprot.org/uniref/search?size=1&query=identity%3A1.0
- Human release statistics (v50) GENCODE · June 30, 2026. https://www.gencodegenes.org/human/stats.html
- The 2025 Report on the Human Proteome from the HUPO Human Proteome Project Journal of Proteome Research · 2026. https://pubs.acs.org/doi/full/10.1021/acs.jproteome.5c00759
- The 2025 HUPO HPP report on the human proteome Human Protein Atlas · 2026. https://www.proteinatlas.org/news/2026-02-20/the-2025-hupo-hpp-report-on-the-human-proteome
- Sampling the proteome by emerging single-molecule and mass spectrometry methods Nature Methods · 2023. https://www.nature.com/articles/s41592-023-01802-5
- The One Hour Human Proteome PubMed · 2024. https://pubmed.ncbi.nlm.nih.gov/38579929/
- Proteins by the Numbers National Institute of General Medical Sciences · 2025. https://nigms.nih.gov/biobeat/2025/01/proteins-by-the-numbers/
- Unification of protein abundance datasets yields a quantitative Saccharomyces cerevisiae proteome Cell Systems · 2018. https://www.cell.com/cell-systems/fulltext/S2405-4712(17)30546-X
- How many human proteoforms are there? Nature Chemical Biology · 2018. https://pmc.ncbi.nlm.nih.gov/articles/PMC5837046/
- Intracellular Water Exchange for Measuring the Dry Mass, Water Mass and Changes in Chemical Composition of Living Cells PLOS ONE · 2013. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0067590

Founder and computational chemist, ProteinIQ
Dr. Matic Broz is the founder of ProteinIQ and a computational chemist. He completed a PhD focused on protein structure, molecular dynamics, and neural networks, and writes about structural biology and scientific software.