{"id":"W1615242748","doi":"10.1186/1471-2105-6-236","title":"SIMPROT: Using an empirically determined indel distribution in simulations of protein evolution","year":2005,"lang":"en","type":"article","venue":"BMC Bioinformatics","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":48,"is_retracted":false,"has_abstract":true,"ca_institutions":"University Health Network; University of Toronto; Ontario Institute for Cancer Research","funders":"Canadian Institutes of Health Research; Genome Canada","keywords":"Indel; Alignment-free sequence analysis; Multiple sequence alignment; Sequence (biology); Phylogenetic tree; INDEL Mutation; Sequence alignment; Sequence analysis; Computational biology; Biology; Genetics; Molecular evolution; Protein family; Computer science; Algorithm; Peptide sequence; Gene","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":[0.003274893,0.0007293396,0.0007903606,0.0008938969,0.001050265,0.0008578639,0.00209571,0.001653046,0.003161858],"category_scores_gemma":[0.009755727,0.0006991669,0.0005777662,0.001028131,0.0009261571,0.001302513,0.0007745923,0.001324282,0.0005297521],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001164998,"about_ca_system_score_gemma":0.001022275,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004437762,"about_ca_topic_score_gemma":0.003964669,"domain_scores_codex":[0.9991016,0.0005179996,0.00004250058,0.0001205318,0.0001533063,0.00006401961],"domain_scores_gemma":[0.9935336,0.005036877,0.0003059385,0.000449392,0.0004443172,0.0002299646],"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":[0.0004180952,0.0001312165,0.00627848,0.0001702141,0.00009302932,0.0002754071,0.0001966209,0.9591662,0.003806339,0.01339399,0.003223979,0.01284644],"study_design_scores_gemma":[0.00004664542,0.000026633,0.0002463195,0.000006950349,0.000006667501,0.00004186915,0.000012236,0.9944906,0.001348051,0.003040702,0.0007227259,0.00001060167],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3511471,0.0003233434,0.6250705,0.0005244052,0.0001414196,0.0002740544,0.002058399,0.01392209,0.006538785],"genre_scores_gemma":[0.6443266,0.0002031449,0.3506442,0.0001896873,0.00003546732,0.0006707342,0.001505432,0.001611471,0.0008132565],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004437762,"threshold_uncertainty_score":0.01731944,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02861468696996599,"score_gpt":0.2802764922088181,"score_spread":0.2516618052388521,"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."}}