{"id":"W3160244225","doi":"10.26434/chemrxiv.14045072.v1","title":"A Bag of Tricks for Automated De Novo Design of Molecules with the Desired Properties: Application to EGFR Inhibitor Discovery","year":2021,"lang":"en","type":"preprint","venue":"ChemRxiv","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Division of Chemistry; Office of Naval Research; Ministero dello Sviluppo Economico; Genome Canada; Fundação de Amparo à Pesquisa do Estado de São Paulo; Ontario Ministry of Economic Development and Innovation; Directorate for Mathematical and Physical Sciences; Merck KGaA; Novartis Pharma; Nvidia; Pfizer; National Institutes of Health; National Science Foundation","keywords":"Reinforcement learning; Computer science; Generative grammar; Artificial intelligence; Deep neural networks; Artificial neural network; Machine learning; Generative model; Deep learning","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008580621,0.0002545802,0.0004296,0.000135491,0.00005757543,0.0001746596,0.001220555,0.000126657,6.379634e-7],"category_scores_gemma":[0.0002845497,0.0001789075,0.0001479295,0.0006176354,0.0001004527,0.0002013281,0.000757977,0.0001558232,6.196445e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001415002,"about_ca_system_score_gemma":0.001270191,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003446488,"about_ca_topic_score_gemma":0.000003191177,"domain_scores_codex":[0.9980589,0.0002576673,0.0004374216,0.000624299,0.0003872267,0.0002344817],"domain_scores_gemma":[0.9974927,0.0006260612,0.0004041386,0.0009294137,0.0004832562,0.00006443707],"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":[0.0001772671,0.0001988101,0.00002374442,0.0006449462,0.0001303668,0.000002506732,0.002484922,0.8393793,0.1502198,0.001297679,0.0001673956,0.005273247],"study_design_scores_gemma":[0.000271724,0.00008177406,0.0003110038,0.000269763,0.00003667422,0.000005743889,0.00008665348,0.6214001,0.3765696,0.0007528029,0.0000267941,0.0001873536],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1989801,0.0001787349,0.7987013,0.0004983895,0.00006924354,0.001446587,0.000005867267,0.0001026369,0.00001712748],"genre_scores_gemma":[0.5826749,0.000003904557,0.4165055,0.00008100386,0.00003387982,0.0006466751,0.00001187268,0.00001952381,0.00002274259],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.3836949,"threshold_uncertainty_score":0.7295638,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04152736925203047,"score_gpt":0.2898238147006849,"score_spread":0.2482964454486544,"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."}}