{"id":"W2118290743","doi":"10.1093/bioinformatics/btp366","title":"Automated protein (re)sequencing with MS/MS and a homologous database yields almost full coverage and accuracy","year":2009,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Advanced Proteomics Techniques and Applications","field":"Chemistry","cited_by":40,"is_retracted":false,"has_abstract":true,"ca_institutions":"Bioinformatics Solutions (Canada); University of Waterloo; Western University","funders":"","keywords":"Protein sequencing; Sequence database; Sequence (biology); Proteomics; Protein methods; Computational biology; DNA sequencing; Tandem mass spectrometry; Database; Peptide sequence; Computer science; Sequence analysis; Mass spectrometry; Biology; Genetics; DNA; Chemistry; Gene; Chromatography","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":[0.00007685454,0.0001844449,0.0001698124,0.00003761594,0.0001717616,0.00008154495,0.0001300237,0.0001161066,0.00003520874],"category_scores_gemma":[0.00005638319,0.0001537047,0.00001607248,0.0001068373,0.00007622374,0.0003843558,0.00007708712,0.0002066514,0.000004610388],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005000203,"about_ca_system_score_gemma":0.00005686327,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003034971,"about_ca_topic_score_gemma":0.000008212865,"domain_scores_codex":[0.999206,0.00000374158,0.0002937026,0.0001531622,0.0001147558,0.0002286457],"domain_scores_gemma":[0.9992322,0.00004231776,0.0001995211,0.0003783526,0.00004367085,0.0001039109],"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":[0.0003451012,0.0002451914,0.0003886505,0.001832281,0.00009797791,0.000109191,0.00344218,0.0001958403,0.8967405,0.009030868,0.002342966,0.08522924],"study_design_scores_gemma":[0.002429382,0.0006854239,0.0002131776,0.001039384,0.0001009063,0.001095094,0.0008758601,0.1954833,0.7821543,0.004673874,0.009737452,0.00151187],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9282958,0.0001271778,0.060924,0.000277121,0.000004517914,0.0006016542,0.0001896848,0.0008000779,0.008779992],"genre_scores_gemma":[0.7100922,0.0001928631,0.2891178,0.0002286799,0.0000248871,0.00008467863,0.0001360455,0.00001599548,0.0001069385],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2281938,"threshold_uncertainty_score":0.6267897,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0146431126850085,"score_gpt":0.2623744513277079,"score_spread":0.2477313386426994,"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."}}