{"id":"W3036441178","doi":"10.1016/j.chempr.2020.05.014","title":"Learning to Make Chemical Predictions: The Interplay of Feature Representation, Data, and Machine Learning Methods","year":2020,"lang":"en","type":"article","venue":"Chem","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":118,"is_retracted":false,"has_abstract":false,"ca_institutions":"Queen's University","funders":"National Institute of General Medical Sciences; National Institutes of Health","keywords":"Machine learning; Artificial intelligence; Computer science; Property (philosophy); Chemical space; Feature (linguistics); Representation (politics); Perspective (graphical); Kernel (algebra); Kernel method; Feature learning; Focus (optics); Feature vector; Random forest; Support vector machine; Mathematics; Chemistry; Drug discovery","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":[0.007617007,0.0009775585,0.001675078,0.002236064,0.000604948,0.004443028,0.001811638,0.001943573,0.001292221],"category_scores_gemma":[0.032327,0.0005193115,0.0008413422,0.002130182,0.001777912,0.006229511,0.002154411,0.002936093,0.0005595163],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009349078,"about_ca_system_score_gemma":0.001401996,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00246208,"about_ca_topic_score_gemma":0.002281634,"domain_scores_codex":[0.9970592,0.00153037,0.0001892728,0.0004075303,0.0007099372,0.0001037756],"domain_scores_gemma":[0.9603406,0.03510048,0.001318027,0.00192128,0.001084127,0.0002355001],"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":[0.0002582926,0.000529691,0.01137873,0.0005801198,0.0002990348,0.0001331428,0.0001168772,0.248435,0.003198938,0.04758346,0.007219495,0.6802672],"study_design_scores_gemma":[0.00002006385,0.00005280395,0.0006669462,0.0000440331,0.00003305446,0.00005955304,0.00002786668,0.891285,0.001849192,0.1046522,0.001283894,0.00002533697],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03025825,0.004221349,0.9577586,0.004623685,0.0001321214,0.00006348146,0.0004579064,0.0009705645,0.001513947],"genre_scores_gemma":[0.6473518,0.004582787,0.3438582,0.001051796,0.0007063227,0.0001618749,0.001035324,0.0001889779,0.001063017],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007617007,"threshold_uncertainty_score":0.04028308,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0505661784080373,"score_gpt":0.3798396244581497,"score_spread":0.3292734460501124,"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."}}