{"id":"W4413639891","doi":"10.1101/2025.08.20.25334127","title":"LLM-based data extraction for a large cancer registry, the Ontario Hereditary Cancer Research Network","year":2025,"lang":"en","type":"preprint","venue":"medRxiv","topic":"Gene expression and cancer classification","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Princess Margaret Cancer Centre; Trillium Health Centre; Children's Hospital of Eastern Ontario; London Health Sciences Centre; Mount Sinai Hospital; Hamilton Regional Laboratory Medicine Program; Genome Canada; Hamilton Health Sciences; University Health Network; Hospital for Sick Children; Ontario Institute for Cancer Research","funders":"Hospital for Sick Children; Government of Ontario; Hamilton Health Sciences","keywords":"Cancer; Hereditary Cancer; Cancer registry; Extraction (chemistry); Medicine; Internal medicine; Breast cancer; Chromatography; Chemistry","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"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":[0.01041705,0.0005172291,0.0004793562,0.00488091,0.001734326,0.002021718,0.001134128,0.0003411784,0.003034539],"category_scores_gemma":[0.04222956,0.000433602,0.0007297491,0.006332704,0.0006161372,0.001370564,0.002355496,0.0004173665,0.0009704724],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.01217351,"about_ca_system_score_gemma":0.02112834,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.45596,"about_ca_topic_score_gemma":0.5727546,"domain_scores_codex":[0.994406,0.00126944,0.0008630049,0.001033643,0.002201229,0.0002267166],"domain_scores_gemma":[0.9660655,0.01470501,0.003434576,0.005083573,0.009891084,0.0008203199],"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":[0.001420301,0.0003409031,0.3413399,0.003129142,0.0003651944,0.00195256,0.008190864,0.01034805,0.02759561,0.003618035,0.1040338,0.4976655],"study_design_scores_gemma":[0.0005129253,0.0003590511,0.6660166,0.0006283024,0.0004636904,0.001144066,0.006395422,0.09526736,0.0500774,0.002790611,0.1761077,0.0002368425],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.6814817,0.00213038,0.120423,0.00635584,0.0001655594,0.006638281,0.1481423,0.0178116,0.01685124],"genre_scores_gemma":[0.5884299,0.0009649981,0.2647151,0.0004691485,0.0000615753,0.002603446,0.1347691,0.0007011555,0.007285553],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.54404,"threshold_uncertainty_score":0.906612,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1382310471913104,"score_gpt":0.4311559061885272,"score_spread":0.2929248589972168,"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."}}