{"id":"W4391685687","doi":"10.1038/s41597-023-02872-y","title":"PLAS-20k: Extended Dataset of Protein-Ligand Affinities from MD Simulations for Machine Learning Applications","year":2024,"lang":"en","type":"article","venue":"Scientific Data","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":22,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"Science and Engineering Research Board; International Institute of Information Technology, Hyderabad; Indian Institute of Technology Delhi; Department of Science and Technology, Ministry of Science and Technology, India","keywords":"Binding affinities; Affinities; Drug discovery; Computer science; Docking (animal); Pipeline (software); Artificial intelligence; Machine learning; Computational biology; Chemistry; Stereochemistry; Biology; Biochemistry","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001169347,0.0001213016,0.0001441698,0.0002340353,0.0003529524,0.001010999,0.002325171,0.00003081046,0.00003720347],"category_scores_gemma":[0.0003490412,0.0001160081,0.00003854702,0.0008245812,0.0001789843,0.001399134,0.001439875,0.0001252319,0.000048696],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002195585,"about_ca_system_score_gemma":0.0003052565,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004667524,"about_ca_topic_score_gemma":0.00006340491,"domain_scores_codex":[0.9979993,0.00009817674,0.0003498196,0.0009582756,0.0004012812,0.0001931675],"domain_scores_gemma":[0.996652,0.001221134,0.0001033993,0.001853033,0.0001083159,0.00006210446],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00004411067,0.0005530384,0.00006498442,0.0005897103,0.0002977633,0.00001100433,0.002153397,0.1340292,0.03959219,0.3645521,0.1697293,0.2883832],"study_design_scores_gemma":[0.00009203297,0.0000106305,0.00002843073,0.00003539575,0.00001464471,9.673543e-7,0.00001970228,0.6709683,0.002076079,0.03198009,0.2946791,0.00009466862],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.001699418,0.0007953117,0.865077,0.0002417856,0.0005509027,0.0005799856,0.1308662,0.0001231144,0.00006630524],"genre_scores_gemma":[0.2449325,0.000007718431,0.5464003,0.00003483461,0.0001771364,0.0001589771,0.2073054,0.00002589943,0.0009571573],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.536939,"threshold_uncertainty_score":0.9749082,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0871256239288214,"score_gpt":0.3650666403697066,"score_spread":0.2779410164408852,"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."}}