{"id":"W4402704193","doi":"10.1101/2024.09.12.612771","title":"Escaping the drug-bias trap: using debiasing design to improve interpretability and generalization of drug-target interaction prediction","year":2024,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Interpretability; Computer science; Drug; Machine learning; Artificial intelligence; Virtual screening; Generalizability theory; Computational biology; Drug discovery; Data mining; Bioinformatics; Biology; Pharmacology; Mathematics","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.003162677,0.0004726366,0.0004774273,0.0004717624,0.0002009888,0.0007439428,0.0008013297,0.0001830153,0.000003539956],"category_scores_gemma":[0.0005705822,0.0004403302,0.0001655132,0.0009361164,0.0001035177,0.0005840762,0.001719115,0.000674154,0.000003988707],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000552637,"about_ca_system_score_gemma":0.0006629642,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001383808,"about_ca_topic_score_gemma":0.000003082975,"domain_scores_codex":[0.9958102,0.001135964,0.0008761879,0.001288481,0.0005284609,0.0003607282],"domain_scores_gemma":[0.9970537,0.0005929534,0.0005188099,0.001123499,0.0005498652,0.0001611792],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00007501414,0.0001409397,0.001670535,0.001028005,0.0002885144,0.000009449995,0.001395892,0.415937,0.5758594,0.003006196,0.0001254004,0.000463689],"study_design_scores_gemma":[0.0001135526,0.00002367582,0.005797821,0.0006656617,0.0000800475,9.401716e-8,0.00001318322,0.7495688,0.2430713,0.0002850386,0.00005282729,0.000328073],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4476648,0.0005192519,0.5486577,0.0002149815,0.00202459,0.0006961154,0.00004184452,0.0001791748,0.000001577676],"genre_scores_gemma":[0.7768354,0.00002480087,0.2226661,0.00009669723,0.0002436267,0.00007958397,2.374229e-7,0.00005267755,8.792019e-7],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.3336317,"threshold_uncertainty_score":0.9998049,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03420789716835802,"score_gpt":0.2763697439704992,"score_spread":0.2421618468021412,"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."}}