{"id":"W4416935774","doi":"10.1093/jamiaopen/ooaf134","title":"Biomedical data repositories require governance for artificial intelligence/machine learning applications at every step","year":2025,"lang":"en","type":"article","venue":"JAMIA Open","topic":"Scientific Computing and Data Management","field":"Decision Sciences","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"National Human Genome Research Institute; National Institutes of Health","keywords":"Corporate governance; Data collection; Data governance; Information governance; Clinical governance","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":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication","open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.0066118,0.0001251906,0.0002536905,0.0001190999,0.0009805367,0.001719447,0.005885087,0.00006025625,0.0001341804],"category_scores_gemma":[0.00657412,0.00009818224,0.00005362811,0.001371949,0.0001802474,0.000774058,0.008654849,0.000136803,0.0002521767],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001021627,"about_ca_system_score_gemma":0.0001606764,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000300698,"about_ca_topic_score_gemma":0.0002995789,"domain_scores_codex":[0.9966357,0.0001164876,0.0008296591,0.001276892,0.0008775296,0.0002636815],"domain_scores_gemma":[0.9951231,0.001459541,0.0003213197,0.002760626,0.0002505348,0.00008484247],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00007434032,0.00008459666,0.0007017726,0.00001361959,0.00002513108,0.000002458391,0.00008226103,0.0001706872,0.00005896741,0.06219716,0.3069409,0.6296481],"study_design_scores_gemma":[0.00007779355,0.00002474503,0.0004587612,0.00002735803,0.00001458832,0.000001281739,0.0004949635,0.1167166,0.00009640832,0.01834971,0.8636349,0.0001028353],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001154099,0.0002817373,0.9776511,0.007377959,0.002619341,0.001261874,0.000509401,0.00008046351,0.00906405],"genre_scores_gemma":[0.5812216,0.00007615552,0.08461132,0.001690169,0.001416436,0.0006494955,0.002753507,0.00004283727,0.3275385],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8930398,"threshold_uncertainty_score":0.9994935,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2400211051165123,"score_gpt":0.4668468828146544,"score_spread":0.2268257776981421,"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."}}