{"id":"W1983546999","doi":"10.1186/1471-2105-9-s12-s7","title":"Extracting transcription factor binding sites from unaligned gene sequences with statistical models","year":2008,"lang":"en","type":"article","venue":"BMC Bioinformatics","topic":"Genomics and Chromatin Dynamics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Alpha Technologies (Canada)","funders":"National Science Council","keywords":"DNA binding site; Chromatin immunoprecipitation; Computational biology; DNA microarray; Transcription factor; Computer science; False positive paradox; Bayes' theorem; Genetics; Binding site; Biology; Statistical model; Gene; Data mining; Bayesian probability; Gene expression; Artificial intelligence; Promoter","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001038786,0.0009587228,0.0007959673,0.001812851,0.0003658222,0.0005574945,0.00108016,0.0007353857,0.001302329],"category_scores_gemma":[0.005146936,0.0008258636,0.001253636,0.001316245,0.000451858,0.0006235615,0.0005901508,0.001096247,0.0007269602],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007334886,"about_ca_system_score_gemma":0.001850175,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005602872,"about_ca_topic_score_gemma":0.007624027,"domain_scores_codex":[0.9995489,0.0001249101,0.00003220116,0.0001419128,0.00012341,0.00002869742],"domain_scores_gemma":[0.9973158,0.002225508,0.0001466798,0.000107541,0.0001590008,0.00004546187],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002763371,0.0001374769,0.007147067,0.0002227285,0.0001623139,0.0002285656,0.00009930344,0.8507462,0.01593735,0.004270644,0.0007949027,0.1199771],"study_design_scores_gemma":[0.000009961378,0.00001394249,0.0005382646,0.000004799637,0.00001552871,0.00003662987,0.000004998321,0.9932603,0.00212015,0.003652584,0.0003339041,0.000008897223],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02612703,0.00008868775,0.9722404,0.00004089217,0.000004284124,0.00003423668,0.0002835944,0.001048021,0.0001327643],"genre_scores_gemma":[0.1767882,0.0002408089,0.8187459,0.00007511886,0.00002173473,0.0002912259,0.002937841,0.0002551287,0.0006441162],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005602872,"threshold_uncertainty_score":0.01114053,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03288224196300695,"score_gpt":0.2342696851814729,"score_spread":0.2013874432184659,"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."}}