{"id":"W2935957056","doi":"10.1016/bs.mie.2019.03.015","title":"Exploring the sequence, function, and evolutionary space of protein superfamilies using sequence similarity networks and phylogenetic reconstructions","year":2019,"lang":"en","type":"article","venue":"Methods in enzymology on CD-ROM/Methods in enzymology","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":27,"is_retracted":false,"has_abstract":false,"ca_institutions":"Canada's Michael Smith Genome Sciences Centre; Michael Smith Health Research BC; University of British Columbia","funders":"","keywords":"Phylogenetic tree; Computational biology; Biology; Sequence (biology); Protein sequencing; Similarity (geometry); Alignment-free sequence analysis; Function (biology); Phylogenetics; Sequence alignment; Evolutionary biology; Protein function; Sequence space; Protein superfamily; Genetics; Peptide sequence; Gene; Artificial intelligence; Computer science","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":[0.002149153,0.0006870106,0.0006894121,0.005446701,0.001258325,0.001774624,0.0008407063,0.0008483314,0.001998755],"category_scores_gemma":[0.00622412,0.0004709492,0.001296288,0.003159519,0.0006429061,0.002042939,0.001192908,0.001390543,0.0006634624],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008680621,"about_ca_system_score_gemma":0.0008239332,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002810252,"about_ca_topic_score_gemma":0.00434528,"domain_scores_codex":[0.9991598,0.0003660753,0.00004683458,0.0002484636,0.0001214014,0.00005751502],"domain_scores_gemma":[0.9977093,0.001504794,0.0003062376,0.000199118,0.0001318309,0.0001487898],"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":[0.002617652,0.0008196782,0.1611656,0.001527953,0.001753444,0.0009856636,0.002696675,0.3492391,0.1391468,0.04731441,0.002464031,0.290269],"study_design_scores_gemma":[0.00005040933,0.0001309946,0.02699174,0.0001281656,0.0002274259,0.0004407264,0.001069186,0.9090258,0.00668949,0.05121258,0.003971626,0.00006185604],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.8542912,0.001416325,0.1394458,0.0004588432,0.00001380712,0.00005863483,0.00148992,0.0007548924,0.002070572],"genre_scores_gemma":[0.9218811,0.0007853014,0.07418519,0.0000528827,0.00001351802,0.00005399689,0.002469566,0.0001494307,0.0004090201],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.005446701,"threshold_uncertainty_score":0.01136595,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1068666795368128,"score_gpt":0.3591240988197934,"score_spread":0.2522574192829806,"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."}}