{"id":"W4282940968","doi":"10.1038/s41594-022-00785-9","title":"Screening thousands of transcribed coding and non-coding regions reveals sequence determinants of RNA polymerase II elongation potential","year":2022,"lang":"en","type":"article","venue":"Nature Structural & Molecular Biology","topic":"RNA Research and Splicing","field":"Biochemistry, Genetics and Molecular Biology","cited_by":50,"is_retracted":false,"has_abstract":false,"ca_institutions":"","funders":"National Institute of General Medical Sciences; Canadian Institutes of Health Research","keywords":"RNA polymerase II; Polyadenylation; Biology; Transcription (linguistics); Coding region; RNA; Genetics; Computational biology; RNA splicing; Gene; Cell biology; Gene expression; Promoter","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":[0.0002575814,0.0001780179,0.0003083003,0.0001316683,0.0003265777,0.000007902029,0.000271568,0.0002555928,0.00001570549],"category_scores_gemma":[0.0001051273,0.0001664234,0.0001313515,0.0001867695,0.000195336,0.000007938856,0.000271828,0.0004078819,3.119617e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001380463,"about_ca_system_score_gemma":0.00006152683,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008081231,"about_ca_topic_score_gemma":0.00001027873,"domain_scores_codex":[0.9985503,0.0002038705,0.0003314137,0.0003977687,0.0001853788,0.0003312857],"domain_scores_gemma":[0.9993321,0.00001514416,0.0002162738,0.0002581151,0.00009484191,0.00008355804],"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":[0.0002297284,0.000006904752,0.00201144,0.00003463777,0.00005361957,0.00002213766,0.00006069949,0.00007629885,0.9932671,0.000586354,0.00001862803,0.003632458],"study_design_scores_gemma":[0.0007773554,0.0006828221,0.003421178,0.00002854081,0.00003193556,0.0001596818,0.0001162703,0.0013022,0.9928416,0.0004199865,0.00003030819,0.0001881473],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9913082,0.001451241,0.006565219,0.0001753613,0.0001138796,0.0002240226,0.0001256809,0.000006747624,0.00002966431],"genre_scores_gemma":[0.9978424,0.00008103565,0.001581215,0.0001538797,0.00005245341,0.00001281622,0.0001974007,0.00001896575,0.00005988686],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.006534162,"threshold_uncertainty_score":0.6786549,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009633817608647455,"score_gpt":0.3125940391978635,"score_spread":0.3029602215892161,"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."}}