{"id":"W1981890639","doi":"10.1186/1471-2105-10-206","title":"Representative transcript sets for evaluating a translational initiation sites predictor","year":2009,"lang":"en","type":"article","venue":"BMC Bioinformatics","topic":"Machine Learning in Bioinformatics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary","funders":"Calgary Laboratory Services","keywords":"Benchmark (surveying); Set (abstract data type); Population; Computer science; Data mining; Data set; Expression (computer science); Sequence (biology); Test set; Computational biology; Algorithm; Biology; Machine learning; Artificial intelligence; Genetics","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":[0.00374881,0.001085213,0.0007208796,0.00347876,0.0007979388,0.001077474,0.001350756,0.001049462,0.001347043],"category_scores_gemma":[0.01148118,0.0002681362,0.0009235629,0.003669773,0.0006695118,0.0009665171,0.0009570376,0.000879633,0.0008129161],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00092635,"about_ca_system_score_gemma":0.001558629,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001986562,"about_ca_topic_score_gemma":0.002085975,"domain_scores_codex":[0.9976512,0.0005796275,0.000299104,0.0004097339,0.0009262107,0.0001341094],"domain_scores_gemma":[0.988768,0.006550138,0.0009035523,0.0008961627,0.002554447,0.0003275835],"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":[0.002709844,0.001377208,0.1255165,0.001925197,0.0005978569,0.0008812135,0.0004219434,0.4081754,0.0705333,0.005256532,0.01299935,0.3696057],"study_design_scores_gemma":[0.0001330811,0.001597375,0.02150147,0.00007736874,0.0001923299,0.0006176836,0.0002826374,0.8931029,0.07062212,0.004282627,0.007518944,0.00007138848],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.6340601,0.001083627,0.3406209,0.0002806545,0.0001043525,0.0006363016,0.016479,0.004948176,0.001786866],"genre_scores_gemma":[0.5364089,0.0004801081,0.3923115,0.00008107931,0.00006936496,0.001447698,0.06829531,0.0002562674,0.0006497796],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.00374881,"threshold_uncertainty_score":0.01982582,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04719615713034547,"score_gpt":0.3568267596750517,"score_spread":0.3096306025447063,"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."}}