{"id":"W4293079743","doi":"10.1101/2022.06.07.495155","title":"MARSY: A multitask deep learning framework for prediction of drug combination synergy scores","year":2022,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Mila - Quebec Artificial Intelligence Institute; McGill University","funders":"Natural Sciences and Engineering Research Council of Canada; Institut de Valorisation des Données; Canada First Research Excellence Fund; McGill University; Génome Québec; Government of Canada","keywords":"Machine learning; Python (programming language); Computer science; Artificial intelligence; Encoder; Drug; Multi-task learning; Task (project management); Deep learning; Medicine; Pharmacology","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"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":[0.001640317,0.001561178,0.00141776,0.0008938867,0.0002647079,0.0008826484,0.002106387,0.001757037,0.002887446],"category_scores_gemma":[0.003177733,0.0007503357,0.001391486,0.0005853217,0.0005120126,0.001051186,0.001222602,0.002318841,0.0007771408],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001167184,"about_ca_system_score_gemma":0.001797879,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008972436,"about_ca_topic_score_gemma":0.01165964,"domain_scores_codex":[0.9995309,0.0001412979,0.00002745516,0.0001348998,0.00009079488,0.00007464987],"domain_scores_gemma":[0.9992312,0.0004283678,0.00008543551,0.00004914249,0.0001275958,0.00007830678],"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":[0.0002889296,0.0001916698,0.002391744,0.0001317683,0.000172287,0.0001074653,0.00002469863,0.9306128,0.001862255,0.001942498,0.006043612,0.0562303],"study_design_scores_gemma":[0.000008423535,0.00002326015,0.00006452815,0.000003902717,0.000005367715,0.000003952752,0.000001310004,0.9984542,0.0002924171,0.0009667584,0.0001720926,0.000003672555],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1344998,0.004019798,0.8363395,0.001931354,0.0002523852,0.0002256144,0.006114933,0.01397707,0.002639511],"genre_scores_gemma":[0.7877598,0.0009165364,0.1947907,0.001246338,0.0001915249,0.000594256,0.008545614,0.0003899334,0.005565329],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.008972436,"threshold_uncertainty_score":0.01784045,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01651695736724328,"score_gpt":0.2575314275847669,"score_spread":0.2410144702175236,"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."}}