{"id":"W2116063068","doi":"10.1093/bioinformatics/btl379","title":"General framework for developing and evaluating database scoring algorithms using the TANDEM search engine","year":2006,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Advanced Proteomics Techniques and Applications","field":"Chemistry","cited_by":218,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"National Cancer Institute; U.S. Public Health Service","keywords":"Computer science; Database search engine; Database; Function (biology); Software; Suite; Information retrieval; Matching (statistics); Source code; Data mining; Open source; Search engine; Programming language; Mathematics","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":[0.01630457,0.004227728,0.00238707,0.005702578,0.001588137,0.006931696,0.01036084,0.003550432,0.01269421],"category_scores_gemma":[0.0286655,0.002269382,0.003662998,0.003740079,0.001521574,0.004663006,0.004279596,0.004090745,0.01140829],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001762283,"about_ca_system_score_gemma":0.005254486,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008396735,"about_ca_topic_score_gemma":0.005204038,"domain_scores_codex":[0.9909493,0.002318936,0.001656662,0.0009690038,0.003633905,0.000472117],"domain_scores_gemma":[0.9920678,0.00255886,0.0004119829,0.00134699,0.003298863,0.0003154174],"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":[0.0005494696,0.0007464089,0.004265247,0.00234951,0.0006719183,0.0009243089,0.0005479802,0.131512,0.02335074,0.2283575,0.06046835,0.5462567],"study_design_scores_gemma":[0.000359075,0.0003969174,0.001073577,0.0005707716,0.000202236,0.001024514,0.0001470132,0.7371512,0.02355542,0.1322845,0.102973,0.0002618127],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0004026948,0.0001137325,0.9915736,0.0000755265,0.00002794147,0.0004511624,0.0002429111,0.006498639,0.0006136833],"genre_scores_gemma":[0.006672664,0.0002640108,0.9895452,0.00009671856,0.00004254405,0.0009775357,0.001034417,0.0006205725,0.0007463073],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01630457,"threshold_uncertainty_score":0.08622783,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07749868424509176,"score_gpt":0.3742599619674528,"score_spread":0.296761277722361,"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."}}