{"id":"W4396766076","doi":"10.48550/arxiv.2405.01616","title":"Generative Active Learning for the Search of Small-molecule Protein Binders","year":2024,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Monoclonal and Polyclonal Antibodies Research","field":"Medicine","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Canadian Institutes of Health Research; Alliance de recherche numérique du Canada; Québec Consortium for Drug Discovery; Génome Québec; Nvidia","keywords":"Generative grammar; Generative model; Active learning (machine learning); Molecule; Chemistry; Computer science; Artificial intelligence; Organic chemistry","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001752162,0.001061797,0.001546512,0.0009200362,0.000315485,0.001036877,0.001759919,0.001600964,0.003534668],"category_scores_gemma":[0.004417386,0.0008448908,0.0009688082,0.0007856472,0.001987895,0.00106494,0.001639176,0.002115732,0.0006218129],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001279867,"about_ca_system_score_gemma":0.00111956,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002477106,"about_ca_topic_score_gemma":0.003059153,"domain_scores_codex":[0.9994692,0.0002253969,0.00002723741,0.0000897045,0.0001242059,0.00006429431],"domain_scores_gemma":[0.9972036,0.002323518,0.0001234169,0.0001245192,0.0001361581,0.00008888634],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001005066,0.00005521322,0.0006306751,0.0001567864,0.00007729646,0.00005721807,0.00003761632,0.931809,0.001730817,0.0282029,0.001125156,0.03601677],"study_design_scores_gemma":[0.00001402729,0.00001794792,0.00002588494,0.000006426621,0.00000545022,0.000005520411,0.000002853455,0.9871076,0.0003425213,0.01211382,0.0003545767,0.000003329321],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01754147,0.001480372,0.9756159,0.0005736189,0.00005876087,0.00005031766,0.0001297681,0.001143889,0.003405868],"genre_scores_gemma":[0.7367696,0.001488171,0.2503538,0.0009961623,0.0001885017,0.0004969423,0.0007220971,0.0003361245,0.008648621],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003534668,"threshold_uncertainty_score":0.01182467,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1426960812787556,"score_gpt":0.2646436175865514,"score_spread":0.1219475363077958,"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."}}