{"id":"W4411789579","doi":"10.31219/osf.io/94mex_v1","title":"Natural Language Querying of Biological Databases with Large Language Models","year":2025,"lang":"en","type":"preprint","venue":"","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Bioinformatics Solutions (Canada)","funders":"AstraZeneca","keywords":"Computer science; Database; Natural (archaeology); Natural language; Natural language processing; Geography; Archaeology","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01256674,0.001402811,0.001453998,0.002265097,0.001010223,0.004383065,0.003319756,0.001806052,0.002532886],"category_scores_gemma":[0.06612356,0.0009245314,0.002227337,0.00326733,0.002026083,0.009600217,0.004733614,0.002699654,0.001085528],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002358359,"about_ca_system_score_gemma":0.002678751,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006933055,"about_ca_topic_score_gemma":0.0103716,"domain_scores_codex":[0.9831217,0.008735133,0.00160482,0.001694952,0.00449264,0.0003507665],"domain_scores_gemma":[0.9319214,0.0545189,0.002366066,0.007457247,0.003318469,0.0004179561],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002351413,0.001578854,0.01915712,0.007271505,0.001476532,0.001637567,0.007285892,0.242199,0.08663272,0.1393396,0.03876082,0.452309],"study_design_scores_gemma":[0.0001921024,0.0001824455,0.001373266,0.0001497922,0.0001638655,0.000412788,0.001258076,0.8473251,0.03997007,0.08825514,0.02061639,0.0001009125],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1055231,0.001424272,0.8467249,0.002920295,0.00007002003,0.0008120104,0.005101821,0.03236231,0.005061283],"genre_scores_gemma":[0.376529,0.0007608219,0.6091114,0.0008601766,0.00005556721,0.0005314114,0.009407907,0.001700044,0.001043633],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01256674,"threshold_uncertainty_score":0.06646001,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03145828793273563,"score_gpt":0.322858441744356,"score_spread":0.2914001538116203,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}