{"id":"W3173064752","doi":"10.1101/2021.06.01.446655","title":"Screening thousands of transcribed coding and non-coding regions reveals sequence determinants of RNA polymerase II elongation potential","year":2021,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"RNA Research and Splicing","field":"Biochemistry, Genetics and Molecular Biology","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Canadian Institutes of Health Research; National Institutes of Health; European Molecular Biology Organization; National Science Foundation","keywords":"RNA polymerase II; Polyadenylation; Biology; RNA splicing; Transcription (linguistics); Enhancer; Coding region; RNA; Genetics; Messenger RNA; Gene; Primary transcript; Computational biology; 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":[0.0005921329,0.0003557224,0.0004371387,0.0002776079,0.0002052946,0.0005846466,0.0002931008,0.000353258,0.001160489],"category_scores_gemma":[0.0005571614,0.0002588923,0.0004163028,0.000311344,0.0002510745,0.0002198506,0.000257053,0.0005716069,0.0005699386],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003040626,"about_ca_system_score_gemma":0.0001771391,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004879948,"about_ca_topic_score_gemma":0.001375555,"domain_scores_codex":[0.9995735,0.0000527935,0.00003363985,0.0001261364,0.000166044,0.00004792361],"domain_scores_gemma":[0.9994619,0.0002655803,0.000103335,0.00005421463,0.00005474275,0.00006031053],"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":[0.00007640601,0.00003057647,0.001196477,0.00004333083,0.00001330287,0.00002342549,0.0000108935,0.0005160173,0.9961385,0.00004534273,0.00005901519,0.001846746],"study_design_scores_gemma":[0.00001297573,0.0003491252,0.01766929,0.000008903811,0.00005456603,0.0001260655,0.00004168223,0.01315904,0.9668092,0.0001139008,0.001640533,0.00001479015],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9829912,0.0008215199,0.01211713,0.00008013855,0.00002529369,0.0000376284,0.002660143,0.0002975138,0.0009693296],"genre_scores_gemma":[0.9686728,0.0007748571,0.02208479,0.0002061042,0.00002173547,0.00009498334,0.006069277,0.0001977596,0.001877881],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.001160489,"threshold_uncertainty_score":0.003882289,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01983506028202524,"score_gpt":0.2614347758287761,"score_spread":0.2415997155467508,"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."}}