Skip to content
ToolCargo

A practical MCP workflow

Research UniProt protein records by gene and species with AI and MCP

Find UniProtKB protein-record candidates for an exact gene and organism, inspect a canonical accession and preserve source curation and evidence context. Keep representative or precursor length separate from mature chains and isoforms.

Built for: Research agents identifying source protein records without mixing organisms, accession identities or evidence classes.

What to connect

Create a ToolCargo account and use OAuth or an API key with a supported MCP client. Hosted connectors share your plan’s call quota; connect each required MCP endpoint separately. Review provider permissions before starting.

Client setup and test status · Current plans and limits

Run the workflow

  1. 1. Choose the gene and exact organism

    Supply a literal gene symbol and exact organism taxonomy ID to uniprot_gene_search. Select reviewed, unreviewed or any-status records explicitly. The organism filter does not include all taxonomic descendants.

  2. 2. Review candidates and explicit pages

    Keep canonical accessions, organism identity and returned gene spelling. Continue only with the filter-bound token and unchanged inputs. Start a new search if the source release changes; counts and cursors are not a pinned snapshot.

  3. 3. Inspect an exact primary accession

    Use uniprot_protein_details on one canonical accession. Preserve protein names with their source naming category, reviewed class, protein-existence evidence and inactive states. Do not substitute a secondary identifier, redirect or isoform suffix.

  4. 4. Interpret length and versions carefully

    Keep canonical length, precursor context, entry version, sequence version and source release distinct. UniProt protein-existence evidence does not establish sequence correctness or clinical suitability. Cite the Consortium, CC BY 4.0 and ToolCargo modifications.

A prompt to try

Use ToolCargo UniProt to find reviewed records for gene INS in exact organism taxonomy ID 9606. Inspect the selected canonical accession and report its protein names, source evidence class, canonical length, precursor context and source versions. Explain the difference from mature insulin chains, retain missing values and cite UniProt Consortium with CC BY 4.0 and ToolCargo modifications.

What a useful result looks like

An exact gene/species candidate list and canonical protein record with source identity, evidence, length context and versions.

Know the limits

Selected UniProtKB metadata only. No sequence residues, features, function or disease narratives, structures, bulk streams or medical recommendations. No secondary-accession substitution, automatic redirects, pinned snapshots or inferred biological absence.

Common questions

Does reviewed mean clinical approval?

No. Reviewed describes manual UniProt annotation. Protein-existence evidence is a separate source classification; neither establishes clinical approval, general quality or sequence correctness.

Does a canonical length describe a mature protein chain?

Not necessarily. The canonical record can represent a precursor or representative sequence. Preserve its context and do not assume the length of every processed chain or isoform.

Does a deleted record prove a protein cannot exist?

No. Inactive, deleted, redirected or unavailable accessions describe source record states. They do not establish biological absence.

References and tool documentation

Use the provider’s documentation to check the underlying concepts, and ToolCargo’s references for the exact tools, inputs and limits.

Tool references for this workflow

Continue with the tools

View the main connectorSet up your client

Related workflows