AlphaGenome MCP Server

assess_pathogenicity

Predict a variant's effect size across genomic modalities to prioritize research candidates; returns scores and quantiles, not pathogenic or benign classifications.

Instructions

Predicted effect size of a variant across modalities, for prioritization. The tool name is kept for compatibility: it does NOT classify a variant as pathogenic or benign, and classification is always null.

Returns the strongest effect per scorer (score, calibrated quantile, where it was seen), the largest absolute quantile, and, for a single-nucleotide variant answered from the Atlas, the AlphaGenome Variant Impact (AVI) score. avi_score is null on the live path, because the AVI score is served by the Atlas only.

Source: a single-nucleotide variant is answered from the precomputed AlphaGenome Atlas; an indel or multi-nucleotide variant runs live inference (score_variant). Both return the same scorers in the same shape. Chosen automatically, overridable with source, and always stated in the result.

Results are AlphaGenome model predictions for research prioritization, not clinical classifications: scores and calibrated quantiles are reported as returned, and no pathogenic/benign call is made.

Example: "How large is the predicted effect of chr19:44908684 T>C?"

Input Schema

TableJSON Schema

Name	Required	Description
`alt`	Yes	Alternate allele (A, C, G, T; more than one base for an indel)
`ref`	Yes	Reference allele (A, C, G, T; more than one base for an indel)
`source`	No	Optional: where the answer comes from (default: auto). auto = the precomputed AlphaGenome Atlas for single-nucleotide substitutions, live inference for everything else (indels, multi-nucleotide variants); falls back to live only when the Atlas does not hold the variant. atlas = Atlas only, errors instead of falling back. live = always run the model. The result always states which source answered.
`scorers`	No	Optional: scorer names to use instead of the defaults. Names come from atlas_list_scorers and are the same for both sources, except the AVI scorers, which the Atlas alone serves.
`position`	Yes	Genomic position (1-based, hg38)
`chromosome`	Yes	Chromosome (chr1-chr22, chrX, chrY)
`tissue_type`	No	Optional: keep only the tracks of one tissue or cell type. A name (brain, neuron, blood, liver, heart, lung, kidney) or an ontology CURIE (e.g., UBERON:0000955, CL:0000540). Default: all tissues.

Schema Changelog

Changes observed during successful MCP inspections.

Changed6 schema fields changedv0.3.0Sep 28, 2026

changedInput schema / properties / alt / description

Previous value: -"Alternate allele"New value: +"Alternate allele (A, C, G, T; more than one base for an indel)"

changedInput schema / properties / position / description

Previous value: -"Genomic position (1-based)"New value: +"Genomic position (1-based, hg38)"

changedInput schema / properties / ref / description

Previous value: -"Reference allele"New value: +"Reference allele (A, C, G, T; more than one base for an indel)"

addedInput schema / properties / scorers

Added value: +{
+  "description": "Optional: scorer names to use instead of the defaults. Names come from atlas_list_scorers and are the same for both sources, except the AVI scorers, which the Atlas alone serves.",
+  "items": {
+    "type": "string"
+  },
+  "type": "array"
+}

addedInput schema / properties / source

Added value: +{
+  "description": "Optional: where the answer comes from (default: auto). auto = the precomputed AlphaGenome Atlas for single-nucleotide substitutions, live inference for everything else (indels, multi-nucleotide variants); falls back to live only when the Atlas does not hold the variant. atlas = Atlas only, errors instead of falling back. live = always run the model. The result always states which source answered.",
+  "enum": [
+    "auto",
+    "atlas",
+    "live"
+  ],
+  "type": "string"
+}

changedInput schema / properties / tissue_type / description

Previous value: -"Optional: disease-relevant tissue (default: brain)"New value: +"Optional: keep only the tracks of one tissue or cell type. A name (brain, neuron, blood, liver, heart, lung, kidney) or an ontology CURIE (e.g., UBERON:0000955, CL:0000540). Default: all tissues."

First observedDec 29, 2025

TDQS

A4.3/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden and does so well: it discloses that `classification` is always null, that `avi_score` is null on the live path because only the Atlas serves it, exactly which source answers which variant class, that the chosen source is always echoed in the result, and the shape of the returned per-scorer data. It also sets a clear research-use boundary (no pathogenic/benign call).

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is front-loaded with the disambiguation and return shape, and each paragraph has a distinct job (naming, returns, source routing, caveat). It is slightly long and repeats the no-pathogenic-call point twice (once after the name caveat, again in the closing paragraph), which is minor redundancy rather than padding.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 7-parameter tool with no annotations and no output schema, the description is largely sufficient: it explains the return contents (strongest effect per scorer, largest absolute quantile, AVI) and the source semantics. It stops short of a fully enumerated result structure and does not note any cost or latency difference for the live-inference path, which an agent weighing `source` would benefit from.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema already documents all seven parameters including `source`, `scorers`, and `tissue_type`. The description adds only marginal parameter meaning beyond that (the AVI-scorers-only-in-Atlas caveat, which is already in the schema), so the baseline of 3 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The opening sentence gives a specific verb and resource ('predicted effect size of a variant across modalities') and the second immediately neutralizes the misleading tool name by stating it does NOT classify pathogenicity and that `classification` is always null. It also delineates the two answering paths (SNV via Atlas, indel/MNV via live inference), so an agent can tell what the tool will and won't produce without opening the schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It states the purpose context ('for prioritization') and explains the automatic source selection plus the `source` override, including what happens on fallback. It does not, however, name or exclude the very similar sibling `predict_variant_effect` or `predict_tissue_specific`, so the agent must infer routing between closely related tools.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

assess_pathogenicity

Instructions

Input Schema

Schema Changelog

TDQS

Other Tools