lifesciences-deepagents

Orchestrates life sciences APIs to build knowledge graphs using the Fuzzy-to-Fact protocol, combining MCP tools for entity resolution and curl for edge discovery, then persisting to Graphiti. This skill should be used when the user asks to "build knowledge graphs", "find biological connections", "explore drug repurposing", "validate drug targets", or mentions traversing gene→protein→pathway→drug→disease paths, multi-API orchestration, or graph persistence workflows.

Structured quality review framework for Fuzzy-to-Fact life sciences reports. Evaluates reports against protocol requirements using template-specific criteria, distinguishes presentation failures from protocol failures, and applies consistent standards for paraphrasing vs hallucination. Use when the user asks to 'review this report', 'evaluate report quality', 'check protocol compliance', 'validate CQ report', or similar quality assessment requests.

Queries clinical databases (Open Targets, ClinicalTrials.gov) via MCP tools for target-disease associations, target tractability assessment, and clinical trial discovery. Falls back to curl when MCP is unavailable. This skill should be used when the user asks to "validate drug targets", "find clinical trials", "assess target tractability", "discover disease associations", or mentions Open Targets scores, NCT identifiers, target-disease evidence, druggability assessment, or translational research workflows.

Queries genomic databases (Ensembl, NCBI, HGNC) via MCP tools for gene lookup, variant annotation, orthology, and cross-database ID resolution. Falls back to curl when MCP is unavailable. This skill should be used when the user asks to "annotate variants", "find orthologs", "map gene IDs", "analyze linkage disequilibrium", or mentions gene symbols, ENSG IDs, HGNC identifiers, VEP annotation, LD analysis, HGVS notation, or DNA sequences.

Queries pharmacology databases (ChEMBL, PubChem, IUPHAR, Open Targets) via MCP tools for drug mechanisms, target identification, bioactivity profiling, and indication discovery. Falls back to curl when MCP is unavailable. This skill should be used when the user asks to "find drug mechanisms", "identify drug targets", "analyze bioactivity data", "discover drug indications", or mentions ChEMBL IDs, mechanisms of action, IC50/Ki values, drug-target relationships, or compound similarity searches.

Queries protein databases (UniProt, STRING, BioGRID) via MCP tools for protein lookups, protein-protein interactions, functional enrichment analysis, and cross-database ID mapping. Falls back to curl when MCP is unavailable. This skill should be used when the user asks to "find protein interactions", "analyze interaction networks", "perform GO enrichment", "map protein IDs", or mentions PPI networks, UniProt accessions, STRING scores, BioGRID interactions, or protein ID conversion between databases.

Formats professional reports from Fuzzy-to-Fact pipeline output using domain-specific templates, claim-level evidence grading, and standardized citations. This skill should be used when the user asks to "format a report", "summarize findings", "grade evidence", "write up results", or mentions report templates, confidence scoring, evidence levels, or presentation of pipeline output.

Validates synthetic lethality claims from CRISPR knockout screens using BioGRID ORCS 5-phase workflow. This skill should be used when the user asks to "validate synthetic lethality", "query CRISPR essentiality data", "find gene dependencies", "compare cell line screens", or mentions BioGRID ORCS, gene knockout data, essentiality scores (CERES, MAGeCK, BAGEL), or asks to validate claims from published CRISPR papers.