{"id":"W4410812090","doi":"10.1093/bioinformatics/btaf325","title":"HSDSnake: a user-friendly SnakeMake pipeline for analysis of duplicate genes in eukaryotic genomes","year":2025,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University; Dalhousie University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Pipeline (software); Workflow; Genome; Novelty; Gene duplication; Gene; KEGG; Computational biology; Data mining; Similarity (geometry); Information retrieval; Database; Biology; Genetics; Gene ontology; Artificial intelligence; Programming language","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002028568,0.0001559934,0.0003537867,0.0002815352,0.0000486073,0.00001409303,0.0002150254,0.0001043657,0.000003803881],"category_scores_gemma":[0.00006077734,0.0001418384,0.000202436,0.0005109659,0.00006503575,0.000001358459,0.0001372637,0.00003343996,0.000001371705],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001393344,"about_ca_system_score_gemma":0.00008208582,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001507602,"about_ca_topic_score_gemma":0.0001923097,"domain_scores_codex":[0.9989483,0.00001207819,0.0005692277,0.0001688282,0.00007296016,0.0002285868],"domain_scores_gemma":[0.9992579,0.00003360088,0.0001603547,0.0003845265,0.0001295896,0.000034048],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0009186337,0.0008946548,0.4779064,0.003060855,0.01047262,0.000003044109,0.002058769,0.03743365,0.3033286,0.00961684,0.02137455,0.1329314],"study_design_scores_gemma":[0.005502181,0.0009370397,0.2286835,0.0001222344,0.003685079,0.000004664324,0.002696403,0.2143021,0.1475543,0.0007184424,0.3943001,0.001494006],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9451473,0.006098987,0.04529264,0.000222793,0.0001763061,0.0006623444,0.0003219312,0.000006256923,0.002071424],"genre_scores_gemma":[0.9771137,0.001780087,0.01949659,0.0002803775,0.00004045973,0.00007430654,0.0002349333,0.0000126543,0.0009669342],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3729255,"threshold_uncertainty_score":0.5784003,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01280712218572493,"score_gpt":0.2637024476158325,"score_spread":0.2508953254301076,"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."}}