Awesome De Novo Peptide Sequencing

A comprehensive, curated, and interactive map of the de novo peptide sequencing field. Algorithms, post-processors, downstream applications and adjacent tools, covering both deep-learning and classical approaches. Includes a SQLite database of papers, models, authors, affiliations, and venues, alongside a Quarto-based interactive website with Observable JS visualisations tracking publication impact, journal metrics, and GitHub activity across the field.

🌐 Live site: https://jeroen.vangoey.be/awesome_de_novo_peptide_sequencing/

Scope

The repository tracks the de novo peptide sequencing field broadly. Every paper is classified by:

• kind: one of: algorithm (core sequencer), post-processor (re-ranker / FDR / refinement), downstream-application (uses de novo output for biology, e.g. neoantigen discovery), adjacent (DB search hybrids, glycopeptide tools), review (literature survey), benchmark (evaluation framework / dataset), or meta (everything else: commentaries, theses without a method, …).
• is_deep_learning: TRUE / FALSE, so readers can compare DL-based and classical approaches side by side.
• acquisition_mode: DDA, DIA, both, or not applicable.

What's here

• denovo.db: SQLite database of papers, models, authors, affiliations, cities, countries, and venues. The source of truth.
• denovo.sql: committed SQL dump of denovo.db so diffs are reviewable in git.
• index.qmd + _quarto.yml: the Quarto site (single-page narrative with interactive charts powered by Observable JS).
• build_publication_impact.py: offline OpenAlex refresh for global publication citation counts.
• build_citations.py / build_journal_metrics.py / build_repo_metrics.py: offline refresh scripts for the citation graph, venue metrics, and GitHub activity metrics.
• plots.ipynb: Jupyter notebook for offline exploration / sanity checks (static matplotlib figures, not published).

Contributing

Easiest: open an issue with a link to the paper (DOI, arXiv, bioRxiv, OpenReview, …) and I'll wire it into the database for you. Corrections: wrong author lists, missing affiliations, mis-classified kind / is_deep_learning / acquisition_mode, broken links, you name it, are equally welcome via issue.

Advanced: edit the database directly

If you're comfortable with SQLite, the source of truth is denovo.db and you can edit it with any tool: sqlite3 CLI, DB Browser for SQLite, DataGrip. A new paper typically needs:

• One row in algorithm if the model is new (set name, repository, algorithm_family, short_description, kind, is_deep_learning, acquisition_mode).
• One row in publication (title, publication_date, doi, publisher, url, journal, publication_type).
• One row per author in publication_author with the author_order field set.
• One row in publication_algorithm connecting the new publication to its model(s).
• Affiliations: insert into country → city → affiliation and link each author with author_affiliation (re-use existing rows where possible: author names and (affiliation.name, department) are the natural keys).

Then regenerate the human-readable dump so the diff is reviewable:

sqlite3 denovo.db .dump > denovo.sql
git add denovo.db denovo.sql

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

One-time setup per clone: run git config core.hooksPath .githooks to activate the tracked pre-commit hook that automatically regenerates denovo.sql whenever you stage denovo.db — so the two files can't drift out of sync in a commit.

Preview locally

uv sync
uv run quarto preview

License

Dual-licensed so the data and the code each get the convention of their own community:

• Curated catalog + prose (denovo.db, denovo.sql, index.qmd, README and other docs) → CC BY 4.0. Use it for anything, commercial or otherwise; the only ask is appropriate credit (see Citation).
• Python scripts (build_citations.py, build_journal_metrics.py, build_repo_metrics.py, build_publication_impact.py) → MIT. Standard permissive terms; copyright notice must be preserved when redistributing.

The Zenodo deposit is archived under CC BY 4.0 (the umbrella that covers the bulk of the artefact). The MIT terms on the helper scripts ride along for anyone who wants to lift the scripts into a permissively-licensed downstream tool.

Citation

If you use this catalog, please cite it:

Van Goey, J. Awesome De Novo Peptide Sequencing. Zenodo. https://doi.org/10.5281/zenodo.20825737

Machine-readable metadata lives in CITATION.cff; GitHub renders a "Cite this repository" button in the sidebar that exposes BibTeX / APA / RIS / EndNote conversions.