{"id":"W3004898987","doi":"10.1021/acs.jproteome.9b00555","title":"Repeat-Preserving Decoy Database for False Discovery Rate Estimation in Peptide Identification","year":2020,"lang":"en","type":"article","venue":"Journal of Proteome Research","topic":"Advanced Proteomics Techniques and Applications","field":"Chemistry","cited_by":44,"is_retracted":false,"has_abstract":true,"ca_institutions":"SickKids Foundation; Hospital for Sick Children; University of Toronto; University of Waterloo","funders":"Natural Sciences and Engineering Research Council of Canada; Canada Research Chairs; Canada Foundation for Innovation; Ontario Genomics; Genome Canada","keywords":"Decoy; False discovery rate; De Bruijn sequence; Computer science; Identification (biology); Sequence database; Database search engine; Mascot; Data mining; Database; Mathematics; Biology; Information retrieval; Search engine","routes":{"ca_aff":true,"ca_fund":true,"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":[0.01505651,0.001231793,0.001844402,0.003076374,0.001041626,0.001296386,0.003110213,0.002004026,0.001481162],"category_scores_gemma":[0.03586818,0.0005252663,0.001087185,0.003026937,0.001224944,0.001803805,0.00149071,0.002066214,0.001115487],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009959708,"about_ca_system_score_gemma":0.002188413,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00164462,"about_ca_topic_score_gemma":0.00121721,"domain_scores_codex":[0.9886596,0.00488678,0.0008062961,0.002385666,0.002862199,0.0003994671],"domain_scores_gemma":[0.9742097,0.0164653,0.001900255,0.004236483,0.002895513,0.0002928123],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0027262,0.0006549128,0.01199046,0.001498891,0.001018522,0.001320249,0.0005611337,0.1146255,0.1319213,0.05781933,0.009585901,0.6662777],"study_design_scores_gemma":[0.0001440003,0.0003788483,0.003530738,0.00008702399,0.0001949551,0.001569089,0.00007054801,0.85404,0.1024092,0.02883481,0.008566314,0.0001744932],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.005960745,0.0005692352,0.9919572,0.00006569347,0.0000596151,0.00008782688,0.0001686169,0.0008758119,0.0002552803],"genre_scores_gemma":[0.1240871,0.0005466271,0.872097,0.0002121926,0.00009297172,0.0006593736,0.001106936,0.0002369373,0.000960858],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01505651,"threshold_uncertainty_score":0.07962739,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1032918760521459,"score_gpt":0.4207004083951112,"score_spread":0.3174085323429653,"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."}}