{"id":"W4415610459","doi":"10.1016/j.csbj.2025.10.052","title":"Leveraging AutoML to optimize dataset selection for improved breast cancer variants pathogenicity prediction","year":2025,"lang":"en","type":"article","venue":"Computational and Structural Biotechnology Journal","topic":"AI in cancer detection","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"United Arab Emirates University","keywords":"Interpretability; Pathogenicity; Overfitting; Prioritization; Selection (genetic algorithm); Benchmarking; Feature selection; Ranking (information retrieval); Precision medicine","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":[],"consensus_categories":[],"category_scores_codex":[0.008690257,0.00198571,0.001627609,0.003155456,0.001348046,0.003232171,0.002447594,0.001472092,0.001809351],"category_scores_gemma":[0.02013171,0.0005102796,0.002167773,0.002183824,0.0007199689,0.002357085,0.002727915,0.001964245,0.001635757],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001158667,"about_ca_system_score_gemma":0.003043312,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007615698,"about_ca_topic_score_gemma":0.01513343,"domain_scores_codex":[0.995933,0.001704422,0.0002882459,0.001151008,0.0006197806,0.0003035502],"domain_scores_gemma":[0.9945101,0.003038283,0.0002687049,0.00111156,0.0008260588,0.0002451676],"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":[0.002335778,0.00142633,0.1234274,0.002078763,0.001462314,0.0008736477,0.000562893,0.3071597,0.03026353,0.007519164,0.09137473,0.4315158],"study_design_scores_gemma":[0.0002823983,0.0004762066,0.0103382,0.0001530924,0.0002092133,0.0002573407,0.0002547354,0.9440441,0.01427504,0.01305619,0.01655814,0.0000954115],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5230741,0.006883239,0.3449395,0.004381272,0.0009460028,0.00116644,0.04667129,0.06314016,0.008797926],"genre_scores_gemma":[0.5823284,0.001034721,0.2989813,0.002258206,0.0002289903,0.001097347,0.1095992,0.002523014,0.001948806],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.008690257,"threshold_uncertainty_score":0.045959,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009242350494047672,"score_gpt":0.2672117843807706,"score_spread":0.257969433886723,"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."}}