plant-genomics-mcp
by musharna
README.md
# 🌱 plant-genomics-mcp
> **50 tools** for plant-genomics locus lookup over the Model Context Protocol —
> 28 single-locus + 1 motif lookup + 1 region query + 1 variant annotator + 1 gene-set enrichment + 1 BLAST search + 12 parallel-batch + 5 cross-source synthesis variants.
> Free, public sources: Ensembl Plants, Phytozome BioMart, UniProtKB,
> Europe PMC, QuickGO, Planteome, PlantCyc/PMN, g:Profiler, NCBI BLAST,
> Gramene, JASPAR, KEGG, STRING-DB, ATTED-II, ThaleMine, and BAR (Bio-Analytic Resource for
> Plant Biology).
[](https://pypi.org/project/plant-genomics-mcp/)
[](https://github.com/musharna/plant-genomics-mcp/actions/workflows/test.yml)
[](https://github.com/musharna/plant-genomics-mcp/actions/workflows/docker.yml)
[](https://glama.ai/mcp/servers/musharna/plant-genomics-mcp)
[](https://doi.org/10.5281/zenodo.21636352)
<p align="center">
<img src="examples/assets/cc-demo.gif" alt="Claude Code answering a plant-genomics question live — calling plant-genomics-mcp across Ensembl Plants, UniProt, and Europe PMC and synthesizing the AT1G01010 / NAC1_ARATH gene profile in a single turn" width="780">
</p>
## 📦 Install
```bash
# Zero-install — uv fetches and runs it on demand
claude mcp add plant-genomics --scope local -- uvx plant-genomics-mcp
```
<details>
<summary>Other install paths (pipx, Docker, from source)</summary>
```bash
# pipx — installs the CLI onto your PATH
pipx install plant-genomics-mcp
claude mcp add plant-genomics --scope local -- plant-genomics-mcp
# GHCR Docker image
docker pull ghcr.io/musharna/plant-genomics-mcp:latest
claude mcp add plant-genomics --scope local -- \
docker run --rm -i ghcr.io/musharna/plant-genomics-mcp:latest
# From source
git clone https://github.com/musharna/plant-genomics-mcp.git
cd plant-genomics-mcp
python -m venv .venv && .venv/bin/pip install -e .
claude mcp add plant-genomics --scope local -- "$(pwd)/.venv/bin/plant-genomics-mcp"
```
</details>
## 💬 Try it
Once connected, ask Claude a plain-language question — you don't have to
name any tool or remember the chain:
> **"Tell me everything about the Arabidopsis gene AT1G01010 — its
> function, GO terms, KEGG pathways, protein-interaction partners, and
> recent papers."**
Claude fans out across Ensembl Plants, UniProt, QuickGO, KEGG, STRING-DB,
and Europe PMC in a single turn and hands back one synthesized answer.
Swap in any locus and pass `organism=` for cross-species — e.g. rice
`Os01g0100100` (`oryza_sativa`) — and it routes to the right backends
automatically.
## 🛠️ Tools
**50 tools across 23 backends** — Ensembl Plants, Phytozome BioMart,
UniProtKB, Europe PMC, QuickGO, Planteome, PlantCyc/PMN, g:Profiler,
AlphaFold DB, PDBe, InterPro, JASPAR, PANTHER, OrthoDB, AraGWAS, 1001 Genomes, NCBI BLAST,
Gramene, KEGG, STRING-DB, ATTED-II, ThaleMine, BAR.
28 single-locus + 1 motif lookup + 1 region query + 1 variant annotator + 1 gene-set
enrichment + 1 BLAST search + 12 parallel-batch + 5 cross-source synthesis. Most take a
TAIR-style locus (e.g. `AT1G01010`) plus
optional `organism=` (slug / scientific name / common name / NCBI taxid
— 12-plant curated coverage matrix at the `pgmcp://organisms/coverage`
MCP resource). All publish JSON `outputSchema`, EDAM ontology tags, and
behaviour annotations — every tool is `readOnlyHint` + `openWorldHint`, so
hosts can surface them without a destructive-action confirmation prompt.
<details>
<summary>Full tool matrix</summary>
| # | Category | Tool | What it does |
| --- | ----------------------- | --------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| 1 | Gene metadata (live) | `ensembl_plants_lookup_locus` | Fetches gene record from Ensembl Plants REST (any plant species). |
| 2 | Cross-references (live) | `get_gene_xrefs` | Fetches cross-DB references (UniProt, NCBI Gene, TAIR, GO, …) from Ensembl. |
| 3 | Gene metadata (live) | `phytozome_lookup_locus` | Fetches gene record from Phytozome BioMart (any Phytozome proteome). |
| 4 | Protein (live) | `resolve_locus_to_uniprot` | Resolves a locus to its UniProtKB record (Swiss-Prot preferred, TrEMBL OK). |
| 5 | Literature (live) | `locus_literature` | Searches Europe PMC for papers mentioning the locus (free, no API key). |
| 6 | GO annotations (live) | `locus_go_annotations` | Fetches QuickGO GO annotations (locus → UniProt → QuickGO). |
| 7 | Sequence search (live) | `blast_sequence` | NCBI BLAST URLAPI — async Put/Get polling with progress notifications. |
| 8 | Homology (live) | `gramene_homologs` | Fetches Gramene v69 homology entries (ortholog / paralog) with gene_tree_id. |
| 9 | Pathways (live) | `kegg_pathways` | Fetches KEGG pathway memberships. 7 organisms: Arabidopsis (`ath:`, native AGI), + rice (`osa:`), maize (`zma:`), soybean (`gmx:`), barley (`hvg:`), poplar (`pop:`), brachypodium (`bdi:`) bridged via Ensembl → Entrez ID. |
| 10 | Interactions (live) | `string_interactions` | Fetches STRING-DB first-neighbor interaction partners with per-channel score. |
| 11 | Coexpression (live) | `atted_coexpression` | Fetches ATTED-II Ath-u.c4-0 top-N coexpression neighbors with z-scores. |
| 12 | Curator summary (live) | `bar_gene_summary` | Fetches BAR ThaleMine + GAIA-aliases curator summary for an Arabidopsis locus. |
| 13 | Expression (live) | `bar_efp_expression` | Fetches BAR eFP-Browser expression profile (mean ± SD per tissue) for a locus. |
| 14 | Interactions (live) | `bar_aiv_interactions` | Fetches BAR AIV interaction partners (Arabidopsis + rice) with confidence + papers. |
| 15 | Curator summary (live) | `tair_locus_info` | Silent upgrade — alias of `bar_gene_summary`. MCP tool name preserved for clients. |
| 16 | Metabolism (live) | `plantcyc_locus_info` | Walks gene → enzyme → reactions → PlantCyc/PMN pathways (free BioCyc web-services API). The metabolic-pathway view KEGG/GO lack; found=false for non-enzymatic genes. 11 species have a PGDB. |
| 17 | Sequence (live) | `get_sequence` | Fetches a locus's sequence (genomic / cds / cdna / protein) from Ensembl `/sequence/id` — the fetch half of lookup → fetch → BLAST; feed `sequence` to `blast_sequence`. |
| 18 | Region query (live) | `ensembl_region_query` | Lists gene/transcript/cds/exon features overlapping a genomic interval (chr:start-end) via Ensembl `/overlap/region` — "what's in this QTL interval" without a per-locus lookup. |
| 19 | Enrichment (live) | `go_enrichment` | GO + KEGG over-representation for a gene **list** via g:Profiler g:GOSt — "what is my DE / co-expression set enriched for?" Reports unmapped loci; optional custom background. All 12 organisms. |
| 20 | Plant ontology (live) | `locus_plant_ontology` | Plant Ontology (anatomy / dev-stage) + Trait Ontology annotations for a locus via Planteome (Solr) — the plant-specific ontologies GO doesn't cover. by_ontology rollup; taxon-filtered. Strong for 6 species. |
| 21 | Structure (live) | `alphafold_structure` | AlphaFold DB predicted 3D model for a locus (locus → UniProt → model): global mean pLDDT, per-band confidence, modelled span, and mmCIF / PDB / PAE URLs. found=false when no model is deposited. All 12 organisms. |
| 22 | Structure (live) | `experimental_structures` | PDBe experimentally-solved (X-ray / cryo-EM / NMR) structures for a locus (locus → UniProt): best-first PDB id, chain, method, resolution, coverage, residue span. found=false when none deposited (common for plants). All 12 organisms. |
| 23 | Domains (live) | `interpro_domains` | InterPro domain / family architecture (locus → UniProt): each entry's accession, name, type, source_database (Pfam included), integrated InterPro id, and residue spans, plus a count_by_type rollup. All 12 organisms. |
| 24 | TF motifs (live) | `tf_binding_motifs` | JASPAR curated TF DNA-binding profiles for a locus (locus → UniProt → symbol search, then UniProt-confirmed): matrix id, TF class/family, assay type (SELEX / ChIP-seq / PBM / DAP-seq), IUPAC consensus, PubMed refs, logo URL. Fuzzy name hits for _other_ genes are quarantined in `name_only_matches`. Arabidopsis-heavy coverage. |
| 25 | TF motifs (live) | `jaspar_motif` | One JASPAR profile by matrix id (e.g. `MA0570.1`, or `MA0570` for the newest version) including the raw position-frequency matrix — the drill-down companion to `tf_binding_motifs`. |
| 26 | Interactions (live) | `experimental_interactions` | ThaleMine CURATED EXPERIMENTAL interaction partners (BioGRID / IntAct / PSI-MI) for an Arabidopsis locus — per partner: detection method (two hybrid, pull down, ...), PSI-MI relationship type, physical vs genetic, source DB, PubMed IDs, and an evidence count. The experimental counterpart to `string_interactions` (predicted / text-mined). Arabidopsis only. |
| 27 | Function (live) | `locus_gene_rifs` | ThaleMine curated GeneRIF statements — one-sentence, manually curated descriptions of what the gene does, each tied to a PubMed ID (HY5 has 114). Citable functional context that GO terms and raw abstracts don't provide. Arabidopsis only. |
| 28 | Variation (live) | `locus_variants` | Natural (EVA/dbSNP) variants overlapping a locus's genomic span via Ensembl `/overlap/region` — id, source, consequence class, alleles, clinical significance. variant_count + truncated. All 12 organisms. |
| 29 | Variation (live) | `vep_annotate` | Ensembl VEP consequence prediction for a variant (region + allele, not locus) — most-severe consequence + per-transcript SO terms, IMPACT, SIFT/PolyPhen. All 12 organisms. |
| 30 | Orthology (live) | `panther_family` | PANTHER protein family + subfamily (id + name), GO terms by aspect, protein class, and pathways. found=false when unclassified. All 12 organisms. |
| 31 | Orthology (live) | `orthodb_orthologs` | OrthoDB ortholog group (name, evolutionary rate) + cross-species member genes at the Viridiplantae level. organism_count + truncated. All 12 organisms. |
| 32 | Diversity (live) | `aragwas_associations` | AraGWAS genome-wide association hits per locus — score, MAF, SNP effect, phenotype/study. Arabidopsis-only. |
| 33 | Diversity (live) | `arabidopsis_natural_variation` | 1001 Genomes natural-variation SNP effects across 1135 accessions — chr, position, effect, impact, amino-acid change, transcript + gene span. Arabidopsis-only. |
| 34 | Batch (live) | `batch_*` (twelve variants) | Parallel per-locus fanout for tools 1–6, 8–12, 14. Up to 50 loci per call. |
| 35 | Synthesis (live) | `*_synth` / `consensus_homologs` (four) | Compose 2–5 backends in parallel, return a `SynthesisEnvelope` with per-step status. |
| 36 | Synthesis (live) | `gene_report` | One-shot "tell me about this gene" dossier — annotation + xrefs + protein + domains + GO + KEGG + STRING + literature composed into a rendered Markdown `result.markdown` (+ structured `result.sections`). |
</details>
## ⚡ Quickstart
After install, the simplest call returns the Ensembl Plants record for
NAC001 — the canonical worked example used throughout `examples/`:
```jsonc
// arguments
{ "locus": "AT1G01010" }
// result (truncated)
{
"id": "AT1G01010",
"organism": "arabidopsis_thaliana",
"display_name": "NAC001",
"biotype": "protein_coding",
"seq_region_name": "1",
"start": 3631,
"end": 5899,
"strand": 1,
"assembly_name": "TAIR10",
"description": "NAC domain containing protein 1 ..."
}
```
Cross-species — pass `organism=`:
```jsonc
{ "locus": "Os01g0100100", "organism": "oryza_sativa" }
```
In Claude Code, the same prompt fans out across Ensembl, UniProtKB, and
Europe PMC in a single turn ([animated demo](examples/assets/cc-demo.gif)):
<p align="center">
<img src="examples/assets/cc-demo.png" alt="Claude Code (Opus 4.7) calling plant-genomics-mcp 8 times to return the AT1G01010 / NAC1_ARATH record with Ensembl, UniProt Q0WV96, and the top-3 Europe PMC papers" width="820">
</p>
Full per-tool walkthroughs (with real upstream-API transcripts) live in
[`examples/`](examples/):
| Walkthrough | Coverage |
| ----------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------- |
| [`gene_report_AT1G01010.md`](examples/gene_report_AT1G01010.md) | One-shot Markdown gene dossier — 7 backends composed, with graceful KEGG degradation. |
| [`analyze_locus_AT1G01010.md`](examples/analyze_locus_AT1G01010.md) | Ensembl → xrefs → UniProt → Europe PMC → QuickGO chain (5 tools). |
| [`find_homologs_AT1G01010_NAC_domain.md`](examples/find_homologs_AT1G01010_NAC_domain.md) | BLAST + per-hit UniProt enrichment. |
| [`biological_context_AT1G01010.md`](examples/biological_context_AT1G01010.md) | Gramene + KEGG + UniProt + STRING + ATTED-II (5 tools). |
| [`v0.8_synthesis_walkthrough.md`](examples/v0.8_synthesis_walkthrough.md) | All 4 v0.8 synthesis tools (`*_synth` + `consensus_homologs`) on the same locus. |
| [`cross_organism_walkthrough.md`](examples/cross_organism_walkthrough.md) | v0.9 multi-organism resolver against rice + maize — per-backend routing on PyPI v1.0.4. |
## 📚 Resources & prompts
<details>
<summary>Four read-only MCP resources + three parameterized prompts</summary>
Clients discover them via `resources/list` and `prompts/list`.
**Resources** (`resources/read`):
| URI | What |
| ----------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------ |
| `pgmcp://cache/stats` | Per-backend `TTLCache` rollup — `{hits, misses, size}` for each live backend. |
| `pgmcp://organisms/phytozome` | Slug → Phytozome `organism_id` map. |
| `pgmcp://backends/status` | Per-backend liveness rollup — `name`, `base_url`, `kind`, `subscription_gated`. |
| `pgmcp://organisms/coverage` | Markdown table of all 12 supported plants × 9 ID slots (ncbi_taxid / ensembl / phytozome / string / europe_pmc / kegg / atted / gprofiler / plantcyc). |
**Prompts** (`prompts/get`):
| Name | Required | Optional | Chains |
| -------------------- | ---------- | ------------------------------------------- | ------------------------------------------------------------------------------------ |
| `analyze_locus` | `locus` | `organism` (default `arabidopsis_thaliana`) | Ensembl → xrefs → UniProt → Europe PMC → QuickGO. |
| `find_homologs` | `sequence` | `program` (default `blastp`) | `blast_sequence` → per-hit `resolve_locus_to_uniprot` for UniProt-shaped accessions. |
| `biological_context` | `locus` | `top_n` (default 10) | Gramene → KEGG → UniProt → STRING → ATTED-II. |
</details>
## 🔌 Transports
| Transport | How to launch |
| --------------- | ------------------------------------------------------------------- |
| stdio (default) | `plant-genomics-mcp` (after install) or via Docker above |
| streamable-HTTP | `plant-genomics-mcp-http` — POST JSON-RPC at `http://host:port/mcp` |
The HTTP transport is stateless and emits JSON responses by default —
the right shape for registry indexers and remote hosting.
### Hosted endpoint
A small **personal demo** runs at:
```
https://mjarnoldgt76.tail86d19d.ts.net/mcp
```
Intended for registry indexers, one-off evaluation, and quick
interactive testing — **not for production workloads**. No SLA, no
uptime commitment, URL may change without notice (single laptop on a
residential connection).
```bash
# liveness probe
curl https://mjarnoldgt76.tail86d19d.ts.net/healthz
# {"status":"ok"}
# connect from Claude Code
claude mcp add --transport http plant-genomics-mcp \
https://mjarnoldgt76.tail86d19d.ts.net/mcp
```
For anything beyond casual evaluation, **self-host**. The HTTP transport
is the same binary; self-hosting buys deterministic uptime, your own
bearer-token gate (`PLANT_GENOMICS_MCP_HTTP_TOKEN`), and NCBI BLAST
etiquette under your own contact email.
## ⚙️ Configuration
Stdio needs no configuration. The two env vars that matter:
| Variable | When | Effect |
| ------------------------------- | ------------------- | --------------------------------------------------------------------------------------------------------------------- |
| `PLANT_GENOMICS_MCP_HTTP_TOKEN` | HTTP transport only | Bearer token for `/mcp`; **must be ≥32 chars** or the HTTP server aborts at startup. Generate `openssl rand -hex 32`. |
| `PLANT_GENOMICS_MCP_NCBI_EMAIL` | If you use BLAST | NCBI etiquette contact. Unset → placeholder + per-call warning; NCBI may throttle. |
<details>
<summary>All env vars (HTTP bind, body cap, cache, BLAST concurrency)</summary>
| Variable | Default | Effect |
| -------------------------------------- | ----------------- | ------------------------------------------------------------------ |
| `PLANT_GENOMICS_MCP_HTTP_HOST` | `127.0.0.1` | HTTP bind address. |
| `PLANT_GENOMICS_MCP_HTTP_PORT` | `8765` | HTTP TCP port. |
| `PLANT_GENOMICS_MCP_HTTP_MAX_BODY` | `2097152` (2 MiB) | Reject POSTs with `Content-Length` larger than this. |
| `PLANT_GENOMICS_MCP_HTTP_STATELESS` | `1` | `0` keeps per-client session state (SSE-style). |
| `PLANT_GENOMICS_MCP_HTTP_JSON` | `1` | `0` switches the response shape to streaming SSE events. |
| `PLANT_GENOMICS_MCP_BLAST_CONCURRENCY` | `2` | Max in-flight BLAST searches per process (NCBI per-IP rate limit). |
| `PLANT_GENOMICS_MCP_CACHE_TTL` | `600` | Per-backend TTL+LRU cache entry lifetime, in seconds. 200-only. |
| `PLANT_GENOMICS_MCP_CACHE_SIZE` | `256` | Max entries per backend before LRU eviction. |
| `PLANT_GENOMICS_MCP_CACHE_DISABLED` | unset | Any non-empty value makes every cache a no-op. |
The cache is process-local — restart the server to drop all entries.
Long-running calls (retry storms, multi-second Phytozome BioMart POSTs)
emit MCP `notifications/progress` over the active session; clients opt
in via `progressToken` in the request `_meta`.
</details>
## ⚠️ Error model
<details>
<summary>Wire-prefix taxonomy + batch result shape</summary>
All live tools raise `PlantGenomicsError` subclasses; the MCP SDK
stringifies them into the wire `content` with a `[ClassName]` prefix so
clients can route on failure kind without parsing the message:
| Wire prefix | When |
| ---------------------------- | ------------------------------------------------------------------ |
| `[NotFoundError]` | 404 / empty BioMart row / invalid locus identifier |
| `[RateLimitError]` | 429 retry budget exhausted — back off and retry |
| `[UpstreamUnavailableError]` | 5xx past retry budget — service outage, try a peer backend |
| `[PlantGenomicsError]` | Other (BioMart `Query ERROR:` body, unexpected column count, etc.) |
Batch tools return `{tool, count, results, errors}` where
`results[locus]` is the same shape as the single-locus tool and
`errors[locus]` is the same `[ClassName] message` string. Ensembl's
batch uses the native `POST /lookup/id` endpoint (one HTTP round-trip);
everything else fans out via `asyncio.gather`.
</details>
## 🧪 Development
```bash
.venv/bin/pip install -e '.[dev]' # or: uv sync --extra dev
.venv/bin/pytest -q # unit tests
PLANT_GENOMICS_MCP_LIVE=1 .venv/bin/pytest -q # adds live network probes
PLANT_GENOMICS_MCP_STDIO_SMOKE=1 .venv/bin/pytest -q # adds stdio smoke
.venv/bin/ruff check .
```
With `uv`, pass `--extra dev` — a bare `uv sync` omits (and removes) the test
dependencies. See [CONTRIBUTING.md](CONTRIBUTING.md#dev-setup).
CI runs the unit suite + the stdio smoke on every push/PR (matrix:
Python 3.11, 3.12, 3.13, 3.14 — the full `requires-python` range). The
live-network gate is **not** run in CI to avoid flakes from upstream
availability.
**Scientific validation / drift detection.** `scripts/benchmark_annotations.py`
drives a curated corpus of canonical loci (27, spanning all 12 organisms)
through every backend + synthesis pipeline and compares results to a frozen
baseline, emitting PASS / DRIFT / FAIL plus cross-source consistency
invariants. It's how upstream data drift is caught. A scheduled GitHub Actions
workflow (`.github/workflows/benchmark.yml`) runs it weekly and pages on a
confirmed regression. Operator guide: [`docs/benchmarking.md`](docs/benchmarking.md).
```bash
.venv/bin/python scripts/benchmark_annotations.py # full live sweep (~3-5 min)
```
See [`CHANGELOG.md`](CHANGELOG.md) for release notes, including the
v0.8 → v0.9 `species=`/`organism_id=` → `organism=` migration and the
v1.0.1 HTTP-token enforcement change.
## MCP registry
Listed in the [official MCP registry](https://registry.modelcontextprotocol.io)
under the namespace below (ownership-verification token for `mcp-publisher`):
```
mcp-name: io.github.musharna/plant-genomics-mcp
```
## License
MIT — see [`LICENSE`](LICENSE). Underlying services (Ensembl Plants,
Phytozome, TAIR, PlantCyc, BAR) have their own terms of use; consult
each before bulk querying.
TDQS
A3.7/5.0
Scored across 50 tools
Disambiguation2/5
Several tools occupy overlapping functional territory: tair_locus_info is literally an alias of bar_gene_summary, and locus_variants, arabidopsis_natural_variation, and aragwas_associations all return variant data for a locus. The many batch_* duplicates and synthesis wrappers such as gene_report, analyze_locus_synth, and biological_context_synth further blur boundaries, even though individual descriptions often clarify the intended use case.
Naming Consistency3/5
Names are consistently snake_case and group into recognizable families like locus_*, get_*, and batch_*, but the set mixes verb-first names (get_sequence, resolve_locus_to_uniprot), resource-first names (locus_literature, string_interactions), and source-first names (bar_gene_summary, phytozome_lookup_locus). It is readable and searchable, but there is no single predictable verb_noun convention across the server.
Tool Count1/5
Fifty tools is excessive for one MCP server, especially because 12 are batc_* duplicates of existing single-locus tools and several synthesis wrappers overlap with one another. Even for broad plant genomics, this surface would be far easier to navigate as multiple focused servers or with batc functionality merged into the base tools.
Completeness4/5
The server covers an impressively broad read-only plant genomics workflow: sequence retrieval, BLAST, GO and ontology annotations, pathways, structures, variants, interactions, homology, motifs, literature, and synthesis reports. Minor gaps exist—such as no tissue/developmental expression atlas or synteny browser—but agents can work around them using the existing tools and batch/synthesis endpoints.
Maintenance
ActivityActive
ResponsivenessResponsive