mosaic for drug target intelligence

Pre-clinical drug-target intelligence as MCP tools.

60 deeply-curated oncology targets with compounds, patents, papers, structures, clinical trials and pathways — plus 2,687 co-essentiality partner genes counted on every axis, so synthetic-lethal whitespace rests on measurements rather than on absent data.

12 tools. 5 free: target profiles, compounds, patents, papers, structures. 7 Pro: competitive landscape, pathway context, target validation, clinical pipeline, synthetic-lethal whitespace, resistance-bypass map, target network.

Built around one rule: a value meaning "we did not measure this" is never rendered as a value meaning "this is zero." Counts that hit an ingest cap are reported as floors, genes outside the curated set return "not assessed" rather than zero, and absence claims are refused where the evidence cannot support them.

Bring your own PostgreSQL — no shared instance is provided.