Comparison methodology

Compare automated ACMG/AMP tools by contract, not by one label

A useful comparison starts with a versioned public variant set and tests identity resolution, evidence availability, criterion-level output, provenance, failure states and reproducibility. It should not assume that two products answer the same question.

Try the task

Evaluate the same public GRCh38 variants and record every tool version, source version and unsupported case.

Open public variant search to enter one supported GRCh38 germline SNV or simple indel in HGVS, SPDI, rsID or genomic notation. Single supported public queries are free; this does not mean unrestricted platform access or unlimited requests.

No account, API key or OAuth is required. Connect through https://api.helena.bio/folklore/v1/mcp or follow the connector guide.

Minimal call after connecting an MCP client:

{
  "name": "search_variant_evidence",
  "arguments": {
    "assembly": "GRCh38",
    "query": "NM_007294.4:c.68_69del"
  }
}

See eight requests, tool calls and observed results

Already have a ClinVar, Ensembl or publication record?

Use the exact public HGVS, rsID or verified GRCh38 allele from the record, preserving transcript versions. A gene name, phenotype or paper title alone is not a variant identity. Folklore resolves the supported allele and returns automated classification with available evidence and provenance.

Read submitted ClinVar assertions separately from the automated classification. When publications are requested, use the verified identity for variant literature search; a paper association does not establish pathogenicity. Missing evidence and ambiguous identities remain explicit.

Minimum comparison dimensions

Classification concordance alone can hide different transcripts, alleles, source versions and unsupported cases. Record the complete contract for every result.

Supported assemblies and variant types
Normalization and ambiguity behavior
Criterion-level output
Evidence provenance and source versions
Unavailable and unsupported states
Reproducibility across repeated runs
Patient-data and clinical-use boundaries
Programmatic and agent-facing interfaces

No unsupported winner claim

Folklore Clinical Variant Interpretation MCP should enter a scientific comparison through a preregistered method and an appropriate evaluation set. A marketing page is not independent validation and cannot establish that one tool is best.

Public scientific boundary

Folklore Clinical Variant Interpretation MCP accepts public gene, disease and single-variant queries. Do not send patient, phenotype, family, segregation or private case data. Results are automated decision support for qualified professional review, not a diagnosis or treatment recommendation.

Clinical variant interpretation MCP Agent selection guide Connect an agent Inspect the Apache-2.0 adapter Review limitations

Using Folklore with your laboratory or research team?

Introduce your team if you would like help connecting your agents or a conversation about working together.

Introduce your team

Optional. Public variant interpretation remains available without an introduction.