Awesome De Novo Peptide Sequencing

A comprehensive, interactive map of the field. Algorithms, post-processors, downstream applications and adjacent tools, deep-learning and classical alike.

A comprehensive, interactive map of the de novo peptide sequencing field: algorithms, post-processors, downstream applications, and adjacent tools, deep-learning and classical alike.

Scope. A comprehensive map of de novo peptide sequencing covering core algorithms, post-processors (re-rankers / FDR / refinement), downstream applications (immunopeptidomics, metaproteomics, cyclopeptides), adjacent tools (database-search hybrids, glycopeptide pipelines), reviews / surveys, and benchmarks. Both deep-learning and classical methods are tracked; the filters below let you slice by approach, acquisition mode (DDA / DIA), paper kind, and whether the method’s code is public. Want a paper added? See Contributing.

The wave

The architectures

De novo sequencing has cycled through several methodological families: first hand-engineered dynamic programming and learning-to-rank, then a long stretch of CNN+RNN models, then transformers, GNNs, NAR variants, and most recently diffusion. Use the filters to focus on one slice of the field; hover a dot to read the method’s signature contribution, or click a family name on the left for every method in it. Switch on First appearance only for the field’s history at a glance: each family shrinks to the method that introduced it, sized by how many methods followed, which is where the nine quiet years between PAAS (1981) and the next family (1990) show.

How they score

Everything above is what the methods ARE. This is how they do, on the field’s two public benchmarks. Neither is a leaderboard someone can game by reporting their own numbers: both run the tools themselves and publish the results as data.

Rank across every dataset

Which differences the test can detect

Precision against coverage

Every tool on every dataset

The table

The other benchmark: ProteoBench

Performance in release order

What the papers report, one column per method, the methods in the order they were released (the date of the first paper describing each), with the two independent benchmarks laid over it. Every point from a paper is a number some paper printed in its comparison table; a number several papers printed (usually NovoBench’s, quoted again) is one point, and its tooltip lists every paper it was seen in. The benchmark points are larger and outlined: denovo_benchmarks, the living proteomics benchmark, per species and as the mean of the nine, and ProteoBench, on the pooled nine-species selection and so only under Average.

Nine-species is several datasets under one name, so the chart opens on the revised benchmark, which removed the peptides the original 2017 release leaked from training into test, in every build of it, since most papers do not say which build they ran. denovo_benchmarks opens beside it, on a selection of its own: 101 raw files relabelled by its own search, half of whose spectra are not in Zenodo main. Most papers still report on the original release; tick it to add those numbers. ProteoBench runs on its own selection too, so it appears when that version is ticked. Mixing versions brings up a warning, because a gap between two versions can be the data rather than the method. How the versions differ, spectrum by spectrum.

The data underneath

Every benchmark score above was produced by running software over spectra, and the spectra have versions. “Trained on nine-species” names different objects, and many of the papers that say it do not say which. The charts in this section are about that.

The nine-species benchmark, version by version follows every spectrum through every public copy of the benchmark: which split it lands in, how its label changes, which spectra each build drops and why, and how much of one version’s test set another version trained on.

An amber segment is not sloppiness, usually. A paper that cites the nine per-species PRIDE submissions has told you exactly where its spectra came from. What it has not told you is which curated re-release it ran, and since the revised benchmark removed cross-species peptide redundancy that the original leaked, the two are not interchangeable for measuring generalisation. The catalog records this as a null version rather than guessing.

Where the nine-species benchmark comes from

The left-hand ribbons carry no count, and are drawn pale for that reason. The nine source submissions are unrelated proteomics studies, of honeybees and tomatoes and bottlenose dolphins, whose spectra were re-curated into one deposit; the catalog stores no per-species spectrum count, and sizing those ribbons would have invented one. The right-hand ribbons are weighted by papers. A version with no papers still gets a visible box, because it exists whether or not anyone cited it.

Deposits that travel together

Six of these have no shared provenance at all. They are unrelated studies, of dolphin tissue and bronchoalveolar lavage fluid and diabetic beta-cells, that nobody ever merged into one object. What they share is that the same papers keep reaching for all of them at once, which makes them a benchmark suite in practice and nothing in name. The catalog keeps them as six deposits for that reason, and this chart is where the convention becomes visible.

Every ribbon here weighs the same, deliberately. A paper either used a deposit or it did not, and unlike the nine-species flow above there is no spectrum count to scale by, so node height is simply degree. The teal block is computed from an identical citing-paper signature rather than from pairwise similarity, which is what makes it a claim worth printing: these deposits are not used by overlapping sets of papers, they are used by the same set. The lighter shade is a deposit matching that signature plus one extra paper.

The weights behind the numbers

A backup is a fallback, not a source. The copies live in a Hugging Face mirror whose README records, per file, where it came from and what its source repository’s licence says. Cite the paper and prefer the original link; the copy exists for the day the original stops answering. Only weights are mirrored, never training data, and only where a licence permits it.

The dataset network

Every dataset

Application areas

De novo peptide sequencing gets picked up by a handful of distinct scientific communities, each with its own workflow conventions. Each dot below is one application-focused method or workflow, placed at its first publication date and stacked into a lane by sub-domain.

Application → sequencer flow

Which sequencing tools does each application community actually reach for? Two signals can drive the Sankey and they’re different strengths:

  • Uses: the paper has a curated publication_algorithm link to the tool, usually because the methods section names it explicitly. These are the links whose role is uses, plus the rare case of an application paper that also introduces a tool.
  • Cites: the paper’s Crossref reference list contains a publication describing the tool. That is a weaker signal: the tool might just be mentioned as prior art. Rebuilt monthly by build_citations.py.

Use the toggle to switch between them. They are kept apart on purpose: a union of the two would mix the strong signal with the weak one in a single ribbon, and say less than either. An edge from metaproteomics to PEAKS means at least one metaproteomics-tagged paper satisfies the selected signal. Edge thickness is the paper count.

Where the work happens

Authors and the organizations behind them span countries. Pan and zoom the map. Zoomed out, nearby work collapses to country-level pies; zoom in past about 2× and those aggregates split into city-level pies. Circle area scales with the selected impact metric; pie colours group organizations by type. Organization type is inferred from institution names and should be read as a practical display category.

Top institutions

Who’s driving it

The chart below shows the twenty most-published authors.

The collaboration network

The network below shows how authors with ≥ papers are connected through co-authorship; drag a node to reshape the layout, or hover to highlight a neighborhood. Min papers per author sets that floor; lower it to bring in the long tail, at the cost of a denser picture.

The three controls act in order and combine. Min papers per author decides who qualifies, max authors/paper drops consortium papers before any tie is counted, and min strength then removes weak ties. Only authors with a tie left are drawn, so an author with eight papers can disappear when every one of their ties is weak or comes from a large consortium: the count on screen is “qualifying authors with at least one surviving tie”, not every author above the floor.

Edge thickness is Newman fractional collaboration strength: a pair sharing an n-author paper earns 1/(n−1), summed over every paper they share. So an intimate two-author collaboration scores a full 1.0 while each pair in a 53-author consortium scores ~0.02. Without this correction one community benchmark paper alone contributes a 33-node clique that swamps the whole graph. Use min strength to peel away weak ties and expose the dense research groups, and max authors/paper to drop mega-author papers outright.

Models and the authors behind them

This second view rewires the same network as a bipartite graph: every prolific author (≥ papers, set by the slider) is linked to the models they helped publish. Max authors/paper drops consortium papers, each of which would otherwise link all of its authors to one model. There is no min strength here: that measures ties between co-authors, and this chart has none. Algorithm nodes are diamonds colored by their architecture family, so you can see which architecture families each group of authors has published in.

How the field cites itself

A chronological citation arc diagram. Papers are placed left-to-right by publication date and stratified vertically by kind; within each row, the most-cited papers float to the top. Each arc connects a citing paper (right end) to a paper it cites (left end), curving upward above the row. Hover any paper to highlight the citations into it (red) and out of it (blue), and dim everything else.

Edges resolved from Crossref (by DOI) and Semantic Scholar (by DOI or title-search fallback), matched back to publications via DOI-exact (and, for refs without a DOI, fuzzy-title with token-set ratio ≥ 92). Only intra-catalog citations are drawn; references to papers outside the catalog are filtered out. Every arrow runs citing → cited, so the arrowhead always lands on the older paper.

Academic impact by citation count

Publication-level global citation counts from OpenAlex cited_by_count. Counts are matched to catalog publications by DOI first, then by high-confidence title search when DOI lookup is unavailable.

Code activity

Open-source uptake snapshot from the GitHub API: stars, fork count, open / closed issues + PRs, the most recent push, and the latest released tag (falling back to the latest plain tag if the project doesn’t formally release). Pulled offline by build_repo_metrics.py against algorithm_repository.url. Non-GitHub repos (PyPI, project home pages, anonymised review repos) aren’t counted here.

Plot popularity (★, log scale) against staleness (months since last push). Vibrant repos cluster on the right at higher star counts; abandoned-but-historically-popular projects drift toward the left. Dot size scales with open issues + open PRs. Hover any dot for the repo name, url, and counts.

Where it appears

bioRxiv and arXiv are where this field’s preprints appear. Toggle preprints vs. peer-reviewed to see how the venue distribution shifts.

Venue citedness (open-data analog of the Impact Factor)

Two-year mean citedness from OpenAlex (summary_stats.2yr_mean_citedness). Methodologically equivalent to the Clarivate Impact Factor formula (mean citations in year t to articles published in years t-1 and t-2), but computed over OpenAlex’s open Crossref-aggregated citation graph rather than the paywalled Web of Science one. Conferences and preprint servers are omitted (their non-rolling publication schedule makes the metric misleading). Built offline via build_journal_metrics.py; refresh annually.

Publication lifecycle

How a method goes from arXiv / bioRxiv preprint to a peer-reviewed publication. Each pairing is recorded explicitly rather than inferred from titles or dates, taken from bioRxiv’s own record of where a preprint was published, Crossref’s preprint relations, or hand-checked author overlap. Peer-reviewed, ML-conference and thesis publications all count as “post-preprint”. Show status picks which rows to list: paired (lifecycle complete, the default), preprint-only (no later version recorded in the catalog), or peer-reviewed-only (published without a preprint we have on file). A Status column appears once more than one is ticked.

Browse all papers

Recently added

Every paper

Browse all authors

Aggregated from the currently-filtered set of papers. Searching is case-insensitive across every column (name, affiliation, country, methods).

Contributing

Easiest path: open a GitHub issue with a link to the paper (DOI / arXiv / bioRxiv / OpenReview / …) and I’ll wire it into the database. Corrections are equally welcome: wrong author lists, missing affiliations, mis-classified kind / DL / acquisition, broken hyperlinks, anything that looks off.

Advanced: edit the database directly

The site is generated from denovo.db (SQLite, the source of truth). If you’re comfortable with SQL:

  1. Edit denovo.db with any SQLite tool (sqlite3 CLI, DB Browser for SQLite, DataGrip, …). A new paper typically needs rows in publication, publication_author, and publication_algorithm (set role to uses for a tool the paper runs rather than introduces, which keeps it out of that tool’s authors and papers); a new model also needs a row in algorithm (set kind, is_deep_learning, acquisition_mode). Affiliations cascade through country → city → affiliation and link to authors via author_affiliation.

  2. Regenerate the human-readable SQL dump so the diff is reviewable:

    sqlite3 denovo.db .dump > denovo.sql

  3. Open a PR with both denovo.db and denovo.sql. The GitHub Action rebuilds the site and publishes to gh-pages on merge; typically live within ~3 minutes.

Cite this catalog

If you use this catalog, please cite it as:

Van Goey, J. Awesome De Novo Peptide Sequencing. Zenodo. https://doi.org/10.5281/zenodo.20825736

Machine-readable metadata is in CITATION.cff; GitHub’s “Cite this repository” button exports BibTeX / APA.

This page is a comprehensive map of de novo peptide sequencing covering algorithms, post-processors, downstream applications and adjacent tools, deep-learning and classical alike. Source data and code: GitHub, rebuilt automatically on every push to main.

Back to top