{"id":"W6958487928","doi":"10.6084/m9.figshare.25302906","title":"Additional file 1 of GPAD: a natural language processing-based application to extract the gene-disease association discovery information from OMIM","year":2024,"lang":"en","type":"article","venue":"Figshare","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary","funders":"","keywords":"Dependency (UML); Association (psychology); Natural language; Information extraction; Feature (linguistics)","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.00189287,0.001519747,0.001161651,0.003070915,0.0007057411,0.001720949,0.002132062,0.001156715,0.7738758],"category_scores_gemma":[0.0184996,0.0007948213,0.001130623,0.00370248,0.0004031536,0.001754067,0.001692407,0.001220789,0.1919267],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001068156,"about_ca_system_score_gemma":0.0018118,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00396609,"about_ca_topic_score_gemma":0.006206966,"domain_scores_codex":[0.9993261,0.0001218445,0.0001169162,0.0001918291,0.0001648327,0.00007851905],"domain_scores_gemma":[0.9868973,0.01036543,0.0005650438,0.0006793918,0.001145961,0.0003468869],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0003908066,0.00009485665,0.001877819,0.00322692,0.00006818215,0.000138277,0.00008190899,0.0007012983,0.0005706881,0.001228926,0.9773791,0.01424129],"study_design_scores_gemma":[0.002892451,0.0002619471,0.0125276,0.001514523,0.000244904,0.0007500248,0.0002785847,0.003998166,0.004567946,0.01447493,0.9583049,0.0001840365],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.0001859288,0.000020169,0.001222715,0.00008334703,0.00002730503,0.00007096212,0.9956859,0.001732087,0.0009715208],"genre_scores_gemma":[0.005394772,0.0001277631,0.01050915,0.0004054339,0.00008970071,0.001333194,0.9741344,0.00281256,0.005193094],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7738758,"threshold_uncertainty_score":0.3225387,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.006152937457157603,"score_gpt":0.2413419467694597,"score_spread":0.2351890093123021,"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."}}