{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003432938,0.0003135235,0.0004907168,0.0005658314,0.0003753456,0.0006892479,0.000366549,0.0005003602,0.001456289],"category_scores_gemma":[0.0007620121,0.0004124671,0.0005352272,0.000632913,0.0003450951,0.000255802,0.0002695953,0.0005831812,0.0005911452],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004233723,"about_ca_system_score_gemma":0.000266147,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007156258,"about_ca_topic_score_gemma":0.002427281,"domain_scores_codex":[0.9996154,0.00004253298,0.00003332866,0.0001363835,0.0001120551,0.00006021386],"domain_scores_gemma":[0.9989329,0.0007010814,0.00009283563,0.00008295196,0.0000629524,0.0001272172],"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":[0.0002736845,0.00005740674,0.003847148,0.00005263083,0.00002921025,0.00010684,0.00003789219,0.0005520592,0.9914742,0.00009953165,0.00008160326,0.003387762],"study_design_scores_gemma":[0.00006442942,0.0005346478,0.05785277,0.00001566701,0.0001883246,0.0008568821,0.0001351419,0.01112679,0.9262261,0.0004787458,0.002483568,0.00003674686],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9943025,0.0004743271,0.003474357,0.00004683105,0.0000148978,0.0000176101,0.001063661,0.0001058682,0.0004998617],"genre_scores_gemma":[0.9877986,0.0003519889,0.006673382,0.0001548826,0.00002204003,0.00004546918,0.003869569,0.00007908924,0.001005033],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.001456289,"threshold_uncertainty_score":0.004871786,"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."}}