{"id":"W2107815233","doi":"10.1186/1471-2105-9-101","title":"Prediction of mucin-type O-glycosylation sites in mammalian proteins using the composition of k-spaced amino acid pairs","year":2008,"lang":"en","type":"article","venue":"BMC Bioinformatics","topic":"Glycosylation and Glycoproteins Research","field":"Biochemistry, Genetics and Molecular Biology","cited_by":153,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Program for New Century Excellent Talents in University; China Agricultural University; University of Alberta","keywords":"Glycosylation; Threonine; Computational biology; Serine; Encoding (memory); Binary classification; Support vector machine; Mucin; Computer science; Amino acid; Biology; Bioinformatics; Artificial intelligence; Biochemistry; Phosphorylation","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004776603,0.000503995,0.0003848824,0.0007312232,0.0001840725,0.0003310403,0.0002722782,0.000362486,0.0004859195],"category_scores_gemma":[0.0007530416,0.0001187918,0.0006037114,0.0006265912,0.0001331295,0.0004169269,0.0002748849,0.0004801633,0.0003475421],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002399729,"about_ca_system_score_gemma":0.0004361568,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001306488,"about_ca_topic_score_gemma":0.001454679,"domain_scores_codex":[0.9998528,0.00002483299,0.00001601298,0.0000457197,0.00004396451,0.00001665198],"domain_scores_gemma":[0.9996777,0.0000692926,0.00009074415,0.00002415213,0.0001024239,0.00003581203],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002420564,0.00067694,0.2349666,0.0007370454,0.0004467261,0.0006520445,0.0001209698,0.08993766,0.4462823,0.0007520893,0.005016906,0.2179901],"study_design_scores_gemma":[0.00004962079,0.0003877376,0.05217331,0.00003831777,0.0001119871,0.0004295725,0.00006542298,0.8791041,0.06528798,0.0005656427,0.001753043,0.00003340134],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9448767,0.0008895042,0.05138383,0.00008120889,0.00003404496,0.00004718004,0.001183127,0.0008208416,0.0006836157],"genre_scores_gemma":[0.9178797,0.0005439649,0.0776354,0.00004275772,0.00001549963,0.00003697543,0.00336917,0.00003862595,0.0004378762],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.001306488,"threshold_uncertainty_score":0.002597749,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05543503999968876,"score_gpt":0.2762255201344427,"score_spread":0.2207904801347539,"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."}}