Neoantigen Prediction (bioSkills)

A Claude Code skill that builds a tumor-to-candidate neoantigen pipeline with pVACtools, emphasizing that binding prediction is the easy part and true positives live downstream.


Type	Claude Skill
Supplier	GPTomics bioSkills (community OSS, MIT)
Availability	GA — part of the bioSkills collection
Pricing	Free / OSS (MIT)
Capabilities	Read/Write — Claude runs the skill’s workflow locally (Bash/Python), not as an MCP tool
Verified	works · 2026-07-20
Security	cleared · 2026-07-20 — provenance matches GPTomics/bioSkills, MIT, no advisories, read-only local pVACtools workflow

How to install

bioSkills is not an npm package — skills are plain markdown/code read directly by the agent. Clone the repo, then either run the installer for the whole category or copy the single skill directory.

Claude Code — clone and install via the bundled script:
```
git clone https://github.com/GPTomics/bioSkills
cd bioSkills
./install-claude.sh --categories "immunoinformatics"
```
The installer copies matching skills into ~/.claude/skills/ (default target). Use ./install-claude.sh --list to preview the skills first.
Claude Code / other agents — copy just this one skill:
```
cp -r bioSkills/immunoinformatics/neoantigen-prediction ~/.claude/skills/
```
(run from inside your clone — the previous step left you in bioSkills/; otherwise replace bioSkills/ with the absolute path of your clone). The skill declares its external dependencies (pVACtools, VEP, HLA typers, LOHHLA) in SKILL.md; install them when prompted on first use.

What it does

Runs the full personalized-neoantigen discovery workflow with the pVACtools suite while enforcing the field’s downstream filters:

pVACseq — translates somatic SNVs/indels into mutant peptides via VEP annotation (Wildtype + Frameshift plugins).
pVACfuse — handles fusion-junction neoantigens.
pVACbind — scores arbitrary peptides without a wild-type comparison.
pVACview — manual re-tiering and candidate selection.
Critical filters — HLA loss-of-heterozygosity (LOHHLA, the silent invalidator), clonal cancer cell fraction (CCF), agretopicity/foreignness quality, and expression thresholds.

The skill centers the predicted → presented → immunogenic validation tiers, stressing that most confidently-predicted strong binders are never presented, and most presented peptides never elicit a T-cell response.

Primary use cases: cancer vaccine target nomination, neoantigen ranking, checkpoint-response biomarker discovery.

Notes

Distributed as a SKILL.md (plus reference material) in the bioSkills collection — Claude executes the workflow locally via Bash/Python rather than as an MCP server. The upstream skill front-matter name is bio-immunoinformatics-neoantigen-prediction; if you invoke it as a namespaced plugin command it resolves under the bioSkills plugin, not as a bare /neoantigen-prediction. Binding-affinity detail is delegated to the mhc-binding-prediction skill and ranking to immunogenicity-scoring. pVACtools and its dependencies (VEP, HLA typers) require separate installation. Upstream directory: immunoinformatics/neoantigen-prediction.

Sources

Installed this tool?

Share feedback — install path, OS, errors, workarounds. The form opens with this tool pre-selected and a link back to this page.