{"id":"W4225674135","doi":"10.26434/chemrxiv-2021-9nhm1-v2","title":"Sanitize It Yourself: Web-based molecular sanitization for machine-generated chemical structures","year":2021,"lang":"en","type":"preprint","venue":"ChemRxiv","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Chemical space; Web service; Software; Space (punctuation); World Wide Web; Software engineering; Drug discovery; Bioinformatics; Operating system","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":[0.001867676,0.0008558872,0.0007724896,0.0009666875,0.0005441635,0.001115359,0.001683826,0.001147494,0.01690882],"category_scores_gemma":[0.004253068,0.0004723357,0.0009003124,0.0007426038,0.0007793984,0.001337636,0.001959001,0.001907884,0.00494676],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005542113,"about_ca_system_score_gemma":0.001034878,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000917084,"about_ca_topic_score_gemma":0.001471728,"domain_scores_codex":[0.9994398,0.0001486465,0.00003000713,0.00006637529,0.0002694141,0.00004582432],"domain_scores_gemma":[0.9985135,0.0006470326,0.0001129587,0.000483233,0.0001729338,0.00007029426],"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":[0.002031999,0.001460084,0.003951452,0.0009978475,0.0004496592,0.0008003507,0.000355313,0.2302194,0.06821189,0.05049735,0.07279959,0.568225],"study_design_scores_gemma":[0.0004236133,0.0001769633,0.0003031383,0.00004761094,0.0000426728,0.0001587666,0.00002320992,0.8968251,0.05255204,0.0209239,0.02848676,0.00003622174],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.06182567,0.0008953852,0.7900591,0.0007812803,0.0003356417,0.0005384285,0.002050052,0.1280332,0.01548117],"genre_scores_gemma":[0.3075752,0.0010627,0.6595425,0.0008625429,0.000125587,0.0009851462,0.006365451,0.01282977,0.0106512],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01690882,"threshold_uncertainty_score":0.05656564,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02913454245077354,"score_gpt":0.305880504540455,"score_spread":0.2767459620896814,"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."}}