SquidBricks: Open-Access Biomedical Literature Screening
The notebook is designed for semi-automated biomedical literature screening. It starts with a PubMed Central (PMC) query, PMC URL, or a list of PMCIDs/PMIDs. It then downloads open-access BioC XML files, parses the full text into passages, filters passages by section and keyword terms, ranks the most relevant text chunks, and uses an LLM classifier to identify relevant articles.