Skip to content
Kitbay
Skill

mining-pubmed-literature

Apache-2.0Updated on Nov 19, 20265,828GitHub stars

A skill from the OpenMed project that teaches an agent to gather biomedical literature from PubMed and PubMed Central through NCBI's E-utilities.

What the skill teaches

A skill from the OpenMed project that teaches an agent to gather biomedical literature from PubMed and PubMed Central through NCBI's E-utilities. It explains the two-step pattern of searching for PMIDs and then fetching or summarising the records, with Python examples that chain the calls through the Entrez history server, use MeSH terms and date limits, and respect the rate limits (three requests per second, ten with a free API key). It is meant for collecting citations and abstracts for a drug, gene or condition, or building a text corpus. It is designed to pair with OpenMed's biomedical entity extraction, which can supply search terms or process the fetched abstracts.

Capabilities

  • Search PubMed for PMIDs with ESearch
  • Fetch abstracts and records with EFetch or ESummary
  • Use MeSH terms and date ranges for reproducible searches
  • Collect abstracts as a corpus for entity extraction