{"id":"W7108685807","doi":"10.5376/cmb.2025.15.0016","title":"Large Language Models for Biological Knowledge Extraction","year":2025,"lang":"","type":"article","venue":"Computational Molecular Biology","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Process (computing); Information extraction; Knowledge representation and reasoning; Knowledge extraction; Relation (database); Domain (mathematical analysis); Relationship extraction; Event (particle physics); Domain knowledge","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0005297104,0.0004348276,0.0004990117,0.0002165801,0.0002851844,0.00004390157,0.0004675032,0.001219321,0.00002964805],"category_scores_gemma":[0.0006035275,0.0004129334,0.0004126593,0.0003021816,0.000510636,0.000005129361,0.0003803989,0.0002777537,0.00002188029],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006177304,"about_ca_system_score_gemma":0.0005563813,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000005380935,"about_ca_topic_score_gemma":0.00001155681,"domain_scores_codex":[0.9971138,0.0003779356,0.0006125983,0.001084624,0.00009222439,0.0007188439],"domain_scores_gemma":[0.9985954,0.0003263786,0.0001886123,0.0003890203,0.0003611101,0.0001395091],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.001057375,0.001712577,0.0006077202,0.0003379099,0.00136336,0.00002624073,0.0004268901,0.006220924,0.243152,0.5570426,0.009509543,0.1785429],"study_design_scores_gemma":[0.006909879,0.002462338,0.0006846365,0.000196029,0.0002691171,0.0000616454,0.001063313,0.1180552,0.0171356,0.5986346,0.2530983,0.001429342],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04798767,0.02385363,0.9212584,0.001036045,0.001193886,0.0005911732,0.0003601998,0.00005776963,0.003661236],"genre_scores_gemma":[0.9695891,0.0001746136,0.02501321,0.001443365,0.0002934517,0.0001627263,0.002367978,0.00002822284,0.0009273014],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9216015,"threshold_uncertainty_score":0.9998323,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02436661784923981,"score_gpt":0.3726838860090235,"score_spread":0.3483172681597836,"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."}}