{"id":"W4399390779","doi":"10.1038/s41597-024-03443-5","title":"The Pan-Canadian Chemical Library: A Mechanism to Open Academic Chemistry to High-Throughput Virtual Screening","year":2024,"lang":"en","type":"article","venue":"Scientific Data","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Winnipeg; University of Alberta; Structural Genomics Consortium; University of Toronto","funders":"National Institute of General Medical Sciences; Genentech; European Federation of Pharmaceutical Industries and Associations; University of Toronto; Ontario Genomics Institute; Merck KGaA; Alliance de recherche numérique du Canada; National Institutes of Health; Ontario Genomics; Genome Canada; McGill University; Bayer; Pfizer; Bristol-Myers Squibb","keywords":"Chemical space; Drug discovery; Workflow; Virtual screening; Computer science; Chemical library; Data science; Cheminformatics; Chemical database; Process (computing); Mechanism (biology); Chemistry; Chemical biology; Resource (disambiguation); Nanotechnology; Small molecule; Database; Physics; Materials science; Computational chemistry; Organic chemistry","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[{"model":"gemma","categories":["open_science"],"domain":null,"study_design":"not_applicable","genre":"methods","about_ca_system":true,"about_ca_topic":false,"confidence":"medium","status":"direct model label, unvalidated"},{"model":"gpt","categories":[],"domain":null,"study_design":"not_applicable","genre":"dataset","about_ca_system":true,"about_ca_topic":true,"confidence":"low","status":"direct model label, unvalidated"}],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication","open_science","insufficient_payload"],"consensus_categories":["open_science"],"category_scores_codex":[0.00293681,0.0002013394,0.0001656463,0.0001659364,0.000640145,0.008770937,0.01697648,0.00008576721,0.0001002739],"category_scores_gemma":[0.0005309423,0.0001655137,0.00004083045,0.002422755,0.0001172459,0.003298231,0.01844238,0.000411926,0.0008688744],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001088179,"about_ca_system_score_gemma":0.001624971,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001305261,"about_ca_topic_score_gemma":0.0003613512,"domain_scores_codex":[0.9963288,0.000118273,0.000336323,0.001843819,0.0007133027,0.0006595083],"domain_scores_gemma":[0.9957167,0.0005191669,0.00003791268,0.003037594,0.00004774689,0.00064086],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000008673074,0.00001148264,0.000001876296,0.00001204774,0.00002967865,0.0000360156,0.0005467653,0.000554123,0.01118503,0.232514,0.527517,0.2275833],"study_design_scores_gemma":[0.00008850065,0.0000151503,0.00005182035,0.0001314811,0.000008671461,0.00002200989,0.00007974558,0.1764573,0.01977137,0.04400361,0.7590272,0.0003431007],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01261392,0.0002387344,0.9222267,0.05316851,0.005261929,0.0009011605,0.003074517,0.0003887428,0.002125855],"genre_scores_gemma":[0.2975014,0.0000110703,0.6698908,0.004929002,0.0009479158,0.0001483044,0.002484648,0.00009471123,0.0239922],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.2848874,"threshold_uncertainty_score":0.999909,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06922418333834772,"score_gpt":0.3512207175254694,"score_spread":0.2819965341871217,"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."}}