{"id":"W4387103580","doi":"10.1101/2023.09.20.558508","title":"An RNA foundation model enables discovery of disease mechanisms and candidate therapeutics","year":2023,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"RNA Research and Splicing","field":"Biochemistry, Genetics and Molecular Biology","cited_by":30,"is_retracted":false,"has_abstract":true,"ca_institutions":"Ontario Genomics","funders":"","keywords":"RNA; RNA splicing; Computational biology; Biology; Polyadenylation; Intron; Exon; Gene; Genetics; Non-coding RNA; Alternative splicing","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0004017488,0.0003029231,0.0002727242,0.0001298715,0.0001110763,0.000187485,0.0003516997,0.0002641595,0.000001865699],"category_scores_gemma":[0.0000714551,0.0003047738,0.00008488727,0.0001171344,0.0001116608,0.00002478039,0.0004153797,0.0002223771,0.000002692462],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003977198,"about_ca_system_score_gemma":0.000688337,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000918742,"about_ca_topic_score_gemma":0.00001451931,"domain_scores_codex":[0.9982647,0.00009261802,0.0002931953,0.0007080904,0.0002762468,0.0003651857],"domain_scores_gemma":[0.9983522,0.00001284407,0.0001962609,0.0009631633,0.0002333926,0.000242112],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0001265432,0.00006034136,0.0008472626,0.0002416156,0.0001166324,0.000006280693,0.000003147459,0.001413156,0.9963724,0.000788265,0.00001962394,0.000004735852],"study_design_scores_gemma":[0.0003991575,0.000152461,0.01045763,0.0001559681,0.0001032381,4.470186e-9,0.000005643513,0.03910423,0.9489444,0.00009234192,0.00009445481,0.0004904117],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9435919,0.00046254,0.05495055,0.00009206843,0.0001542348,0.0003560393,0.0003424691,0.00004894685,0.000001240282],"genre_scores_gemma":[0.9948826,0.001752047,0.002912338,0.00009112484,0.0001394043,0.00007753111,0.000008314069,0.00009648783,0.00004014248],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.05203821,"threshold_uncertainty_score":0.9999405,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02049002437704721,"score_gpt":0.2615424899564699,"score_spread":0.2410524655794227,"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."}}