Study Screening Outputs of "Large language models accelerate dataset screening for meta-analysis in biomedical data repositories"

These are the study screening files from the under-review paper "Large language models accelerate dataset screening for meta-analysis in biomedical data repositories."Human screening decisions from two independent reviewers are provided as CSV files for GEO and ProteomXchange case studies. Screening decisions for the retrospective validation case study (ArrayExpress) originated from Noori et al. (2021): https://doi.org/10.1016/j.nbd.2020.105225Corresponding LLM screening outputs are provided as raw JSONL files, compressed into ZIP archives. Each ZIP archive contains subfolders for each LLM (i.e., gpt-4o-mini, gpt-4.1-mini, o4-mini, and gpt-5-mini).Error analyses are available in corresponding .xlsx files. In each spreadsheet, the right-most "notes" column includes a written note for the cause of misclassification.