{"id":"W2100922775","doi":"10.1093/bioinformatics/btv157","title":"Test set bias affects reproducibility of gene signatures","year":2015,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Gene expression and cancer classification","field":"Biochemistry, Genetics and Molecular Biology","cited_by":123,"is_retracted":false,"has_abstract":true,"ca_institutions":"Princess Margaret Cancer Centre; University of Toronto; University Health Network; Montreal Clinical Research Institute","funders":"National Institute of General Medical Sciences; National Cancer Institute; National Center for Advancing Translational Sciences; BC Cancer Agency","keywords":"Normalization (sociology); Computer science; Data set; Data mining; Set (abstract data type); Test set; Reproducibility; Sample size determination; Population; Test data; Statistics; Pattern recognition (psychology); Artificial intelligence; Mathematics; Medicine","routes":{"ca_aff":true,"ca_fund":true,"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":[],"consensus_categories":[],"category_scores_codex":[0.0006104602,0.00009075185,0.0001044262,0.00003603485,0.00002240475,0.00001147404,0.0001534034,0.0001125406,0.000005557744],"category_scores_gemma":[0.001162637,0.00007455279,0.00004802908,0.0001073532,0.00005217228,0.000005537391,0.00006061255,0.00004728872,0.00001053879],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001196205,"about_ca_system_score_gemma":0.0001305274,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000003134592,"about_ca_topic_score_gemma":0.000002446749,"domain_scores_codex":[0.9992481,0.00002776577,0.0002319678,0.000218386,0.0001617081,0.0001120167],"domain_scores_gemma":[0.9985875,0.00001275206,0.0001746306,0.0009924986,0.0001434232,0.00008914548],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00007227995,0.0001088631,0.008653097,0.0001103799,0.00002345729,3.680587e-7,0.0005649742,0.0002603421,0.8556376,0.00006320964,0.1287787,0.005726838],"study_design_scores_gemma":[0.0003932392,0.0002566707,0.003687503,0.00001147529,0.000009390277,0.000003911169,0.0002896265,0.0007315241,0.9509135,0.00005292802,0.04352804,0.0001221983],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9791114,0.001098194,0.005213644,0.0002653419,0.0004636536,0.0004169295,0.00008845135,0.00003788374,0.0133045],"genre_scores_gemma":[0.9933034,0.00003899307,0.00581893,0.0001610954,0.0001044808,0.000008967813,0.0001526745,0.000007732556,0.0004037753],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.09527595,"threshold_uncertainty_score":0.3040175,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0516341996318329,"score_gpt":0.2907679073513346,"score_spread":0.2391337077195017,"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."}}