{"id":"W2805247015","doi":"10.1101/333914","title":"Pan-cancer machine learning predictors of primary site of origin and molecular subtype","year":2018,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Cancer Diagnosis and Treatment","field":"Medicine","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Cancer Institute; BC Cancer Agency; National Institutes of Health; Cancer Research UK","keywords":"Cancer; Precision medicine; Identification (biology); Personalized medicine; Primary tumor; Primary cancer; Targeted therapy; Genome","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"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":[0.0002899943,0.000423365,0.001012482,0.0002412421,0.00004503612,0.00002233043,0.000127128,0.0003224682,0.00009592433],"category_scores_gemma":[0.00009602509,0.0003957208,0.0001779737,0.0002950021,0.0002295762,0.00004879465,0.0003071087,0.0004604646,0.000006092965],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002912673,"about_ca_system_score_gemma":0.000697905,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0006286154,"about_ca_topic_score_gemma":0.000008509945,"domain_scores_codex":[0.9979538,0.00007490481,0.000554622,0.0006850522,0.0004271404,0.000304529],"domain_scores_gemma":[0.9978975,0.00005732069,0.0006034534,0.0007359785,0.0005002918,0.0002054755],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0001548384,0.0002623673,0.6776768,0.001638801,0.0006617254,0.0000444208,0.00001826221,0.00001654571,0.3193457,0.00004291671,0.0001273329,0.00001033483],"study_design_scores_gemma":[0.001301579,0.0003628273,0.6820644,0.001609697,0.0008840167,3.083947e-8,7.977529e-7,0.0001000846,0.3094637,6.27145e-7,0.003922755,0.0002895422],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9811316,0.01699407,0.00006843378,0.0001499988,0.0004583251,0.000718437,0.0003648789,0.00009189872,0.00002234258],"genre_scores_gemma":[0.9918997,0.00617754,0.001299155,0.0001011223,0.0002661426,0.0001454331,0.00000220771,0.0001009357,0.000007751951],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01081653,"threshold_uncertainty_score":0.9998494,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01306591959602068,"score_gpt":0.2387885973713888,"score_spread":0.2257226777753681,"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."}}