{"id":"W3172130576","doi":"10.1093/sysbio/syab044","title":"A New Pipeline for Removing Paralogs in Target Enrichment Data","year":2021,"lang":"en","type":"article","venue":"Systematic Biology","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":55,"is_retracted":false,"has_abstract":true,"ca_institutions":"Health Sciences Centre; University of Calgary","funders":"","keywords":"Biology; Phylogenetic tree; Pipeline (software); Phylogenomics; Evolutionary biology; Computational biology; Divergence (linguistics); Tree (set theory); Sequence (biology); Phylogenetics; Genetics; Gene; Computer science; Clade; 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":[0.003084894,0.002013119,0.001271453,0.002087682,0.001565377,0.001607561,0.002260709,0.001092624,0.007828983],"category_scores_gemma":[0.004986695,0.001516617,0.002278239,0.001349271,0.000590422,0.002506721,0.003127941,0.003697768,0.007774159],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008120091,"about_ca_system_score_gemma":0.002138849,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003616747,"about_ca_topic_score_gemma":0.004531648,"domain_scores_codex":[0.9974959,0.0001567861,0.0002583535,0.000964944,0.000900181,0.0002237459],"domain_scores_gemma":[0.9976009,0.0006546429,0.000203071,0.000658321,0.000696139,0.0001868488],"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":[0.001263194,0.0004867596,0.01460238,0.001627255,0.000430592,0.0008116852,0.001225778,0.006232375,0.4298182,0.004251677,0.08239084,0.4568592],"study_design_scores_gemma":[0.0004894985,0.0007015542,0.03188644,0.00017799,0.0004692983,0.002084325,0.0003857211,0.2169098,0.4309168,0.01371157,0.3017395,0.0005275129],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02449011,0.0004770726,0.8552529,0.0004067759,0.0003924962,0.0006997099,0.0144533,0.1009128,0.002914881],"genre_scores_gemma":[0.04633902,0.0003258212,0.8894709,0.0007501514,0.00009802259,0.0009993786,0.04918462,0.007093897,0.005738171],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007828983,"threshold_uncertainty_score":0.02619058,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03497564237824111,"score_gpt":0.2969124139618574,"score_spread":0.2619367715836162,"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."}}