{"id":"W2945671922","doi":"10.1038/s41588-019-0411-1","title":"Similarity regression predicts evolution of transcription factor sequence specificity","year":2019,"lang":"en","type":"article","venue":"Nature Genetics","topic":"Genomics and Chromatin Dynamics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":159,"is_retracted":false,"has_abstract":false,"ca_institutions":"Baker Hughes (Canada); Vector Institute; Canadian Institute for Advanced Research; Institute of Cancer Research; Ontario Institute for Cancer Research; Centre for Global Health Research; University of Toronto","funders":"","keywords":"Biology; Sequence motif; Computational biology; Motif (music); Transcription factor; Genetics; Gene; Zinc finger; DNA binding site; Genome; Evolutionary biology; Conserved sequence; Regression; Promoter; Gene expression; Peptide sequence","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":[],"consensus_categories":[],"category_scores_codex":[0.00008824559,0.0001450328,0.0001424989,0.00003474391,0.00003369422,0.000009859062,0.0002130297,0.0005593777,0.00002782589],"category_scores_gemma":[0.00002422092,0.0001335356,0.00009005376,0.00008139743,0.00004592472,0.000002938932,0.00004223477,0.0002418861,0.000005342331],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004576163,"about_ca_system_score_gemma":0.00008775159,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000004552272,"about_ca_topic_score_gemma":0.00001918627,"domain_scores_codex":[0.9990798,0.00003527361,0.0002120062,0.000303101,0.0001987256,0.0001711306],"domain_scores_gemma":[0.9992484,0.000005054494,0.0001269619,0.00041903,0.0001461985,0.0000543231],"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":[0.00004576458,0.00004213617,0.02892851,0.00005807784,0.00001673891,4.076382e-7,0.00003476252,0.0004331862,0.969398,0.00009221022,0.0001961055,0.0007541595],"study_design_scores_gemma":[0.0006844518,0.0004454158,0.1667947,0.00005428588,0.00002594842,0.0000103079,0.00006706407,0.004624072,0.8185231,0.0005913834,0.007881751,0.0002974608],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9947459,0.00222013,0.001548104,0.00003459096,0.00051216,0.0002236476,0.0001744393,0.000009848302,0.0005312095],"genre_scores_gemma":[0.9965788,0.0004793027,0.002364344,0.00004526554,0.0001403839,0.000001997723,0.0001620834,0.00001800614,0.0002098754],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1508748,"threshold_uncertainty_score":0.5445423,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009157682018527683,"score_gpt":0.2409742275952631,"score_spread":0.2318165455767354,"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."}}