Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@GeneBe MCP Is rs41272110 benign? Show ACMG criteria and points."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Give your AI assistant real variant-interpretation data instead of guesses.
Annotate and classify human genetic variants (ACMG, gnomAD, ClinVar, in-silico predictors) from Claude, Cursor, VS Code and any other MCP client, powered by GeneBe.
https://mcp.genebe.net/v1/mcp
Language models are bad at remembering the clinical status of a specific variant. They sound confident and are often wrong. GeneBe MCP lets the assistant look the variant up and answer from data: a transparent, points-based ACMG classification with every criterion explained, together with population frequencies, predictor scores, ClinVar and transcript consequences.
"Is NM_005577.4:c.4195A>C benign?"
"What is chr6-160585140-T-G? Explain why it is classified that way."
"Convert rs41272110 to hg38 and list its consequences on each transcript."
For example, chr6-160585140-T-G (rs41272110, LPA) comes back as Benign, −9 points with criteria BA1 and
BP4, plus the data behind them and a link to the GeneBe page.
Related MCP server: clingen-link
Any notation in: genomic coordinates, HGVS, rsID or SPDI; hg19, hg38 and T2T supported through liftover.
Verdict you can audit: ACMG classification with points and the reasoning for each criterion, not a black-box label.
Evidence in one call: gnomAD frequencies, in-silico predictors, ClinVar, per-transcript consequences (HGVS c./p., impact), gene details with MANE Select and ClinGen gene–disease validity.
Compact answers: large lists are capped, so one call costs roughly 1–3k tokens, and every result also carries structured data that follows the tool's output schema.
Links back to GeneBe, so the human can check the source.
No sign-up, no API key. The server is remote; you only add its URL.
claude mcp add --transport http genebe https://mcp.genebe.net/v1/mcpSettings → Connectors → Add custom connector, then enter https://mcp.genebe.net/v1/mcp.
Open chatgpt.com/plugins, click Add (+) → Add custom MCP, and enter
https://mcp.genebe.net/v1/mcp.
Add to ~/.cursor/mcp.json:
{ "mcpServers": { "genebe": { "url": "https://mcp.genebe.net/v1/mcp" } } }Add to .vscode/mcp.json:
{ "servers": { "genebe": { "type": "http", "url": "https://mcp.genebe.net/v1/mcp" } } }Any client that supports remote MCP over Streamable HTTP works. Try it from a terminal:
curl -s -X POST https://mcp.genebe.net/v1/mcp \
-H 'Content-Type: application/json' -H 'Accept: application/json, text/event-stream' \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"annotate_variant","arguments":{"variant":"chr6-160585140-T-G"}}}'More examples: USAGE.md.
Tool | What it does |
| Start here. Any notation → ACMG verdict, points and criteria, frequencies, predictor scores, ClinVar, consequences, GeneBe links |
| Why each classification criterion applied, in plain text |
| HGVS / rsID / SPDI → hg38 coordinates |
| Per-transcript consequences (HGVS c./p., impact) |
| Gene identifiers, location, MANE Select, ClinGen gene–disease validity, transcripts |
| Positions, intervals or variants between hg19 / hg38 / t2t |
The default genome build is hg38.
Read-only. No tool changes anything, anywhere.
No account, no key needed to use the hosted server.
Single variants only. No patient data, no VCF files, no samples are sent or stored.
Nothing is stored and variant input is not logged by the MCP server.
Transparent classification. Results come from GeneBe's points-based ACMG implementation; the reasoning is returned with the verdict, so you and the assistant can check it.
Open source under the MIT license; you can read exactly what the server does, or run your own copy.
Research use only. This is not a medical device and not a diagnosis. Classifications must be reviewed by a qualified specialist before any clinical use.
Today GeneBe MCP answers questions about single variants. We plan to grow it into a workspace your assistant can use on your own data. None of the items below is available yet, and the order may change.
Sign-in, so the assistant can act on your GeneBe account
Your own samples: view and edit sample details and annotations
Search your cohort: find a variant across your own set of samples
Upload samples from the assistant
Supervise computations: start analyses, follow their progress and read the results
Ideas or requests? Open an issue or write to us via genebe.net.
Results combine data from GeneBe, gnomAD, ClinVar, UniProt, ClinGen and dbNSFP predictors. Please check each source's license terms before redistributing the data.
Want to run your own instance or change the code? See DEVELOPMENT.md. Issues and pull requests are welcome.
Made by the GeneBe team · MIT · Research use only
Annotate and classify one human genetic variant using GeneBe.
Start here for any question about a variant (what is it, is it pathogenic/benign, how common is it). Returns the ACMG verdict with points and criteria, population frequency (gnomAD), predictor scores (REVEL, AlphaMissense, SpliceAI, ...), ClinVar status, transcript consequences, and links to the GeneBe page to cite. Research use only; not a diagnosis.
Inputs: variant (required), genome, transcript, gene_symbol. Returns: gene, dbsnp, links, scores, +11 more
Convert a variant written in any notation into normalized hg38 genomic coordinates.
Use when you only need the chr/pos/ref/alt of a variant (for example to translate HGVS or an rsID), without annotation. Does not classify the variant; use annotate_variant for that.
Inputs: variant (required), input_genome. Returns: error, input, variants
Explain step by step why each classification criterion did or did not apply to a variant.
Use after annotate_variant when the user asks "why was this classified that way?". Returns an ASCII rendering of the decision tree with the evaluated conditions. By default it explains the germline ACMG verdict that annotate_variant returns; the somatic trees 'oncogenicity' (2022) and 'acgs-uk-onco' (2025) can be requested explicitly. Research use only.
Inputs: variant (required), tree, genome, version. Returns: text, tree, variant, version, +2 more
Look up a human gene: identifiers, location, MANE Select transcript, ClinGen gene-disease validity.
Use for questions about a gene rather than a specific variant. Transcripts are listed with MANE Select first and capped; check total_transcripts.
Inputs: gene (required), include_transcripts. Returns: chr, end, link, name, +16 more
List a variant's predicted molecular consequences on every overlapping transcript.
Returns per-transcript HGVS (c. and p.), consequence terms (e.g. missense_variant) and impact, most severe and MANE Select transcripts first. Lightweight: no frequencies or classification.
Inputs: variant (required), genome. Returns: effects, variant, truncated, total_effects
Convert a genomic position, interval or variant between genome builds (hg19, hg38, t2t).
Check status: RESOLVED means one result, AMBIGUOUS means several candidates, anything else
means it could not be mapped (see errors). For variants, ref/alt are rebuilt against the target
reference and warnings such as REF_MISMATCH may apply.
Inputs: query (required), to_genome (required), from_genome (required). Returns: input, errors, status, warnings, +3 more
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 6 tools
Each tool has a distinct primary purpose, and the descriptions explicitly steer agents between similar-sounding tools (e.g. annotate_variant vs get_variant_effects, convert_variant vs liftover_coordinate). Some boundary overlap remains between coordinate normalization/liftover and between annotation/effect retrieval, but it is manageable.
All tool names use consistent snake_case with an action-object structure: get_gene, annotate_variant, convert_variant, get_variant_effects, liftover_coordinate, and explain_acmg_decision_path. The pattern is predictable even where the object is compound.
Six tools are well scoped for a focused human genetics annotation server. Each tool covers a clear, non-redundant step in gene/variant lookup, interpretation, explanation, normalization, effect listing, and liftover.
The surface covers the core workflows: gene lookup, variant annotation/classification, ACMG explanation, variant normalization, transcript effects, and genome-build liftover. Minor gaps exist, such as batch annotation or direct phenotype/disease lookup, but no major workflow dead end is apparent.
MyVariant.info MCP. Aggregated human genetic-variant annotation: dbSNP, ClinVar clinical…
MCP gateway federating 22 biomedical MCP servers behind one endpoint: gnomAD, ClinVar, HPO, VEP.
gnomAD MCP — Broad Institute Genome Aggregation Database (GraphQL).
Germline ACMG/AMP classification, genes, disease associations, literature and API-key case analysis.