{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002306176,0.002148275,0.001605062,0.001948317,0.001433174,0.002444467,0.00379956,0.001363483,0.04588807],"category_scores_gemma":[0.00640111,0.002073059,0.003039164,0.001336776,0.0006760265,0.003082008,0.003435344,0.00295444,0.03291862],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00084107,"about_ca_system_score_gemma":0.001945709,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003174167,"about_ca_topic_score_gemma":0.004963067,"domain_scores_codex":[0.9991586,0.0001211068,0.00007611497,0.0003525813,0.0001983485,0.00009324877],"domain_scores_gemma":[0.9989478,0.0004575797,0.0001049339,0.0002037017,0.0001892476,0.00009682483],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.002143247,0.0001540187,0.008993817,0.004739241,0.001077107,0.0007813214,0.00134581,0.00743577,0.05911174,0.007958041,0.7742198,0.1320401],"study_design_scores_gemma":[0.001227916,0.000439324,0.01991706,0.0006791349,0.0006013283,0.001795263,0.0006739347,0.1235772,0.1201648,0.03672301,0.6931281,0.001072889],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"software","genre_gemma":"software","genre_scores_codex":[0.01033065,0.0005461543,0.2104012,0.0004142065,0.0002867103,0.0003615993,0.07672283,0.6971542,0.003782401],"genre_scores_gemma":[0.06723537,0.0006384883,0.5767841,0.001039607,0.0001091424,0.001704943,0.2000684,0.1415887,0.01083117],"genre_candidate":"software","genre_consensus":"software","teacher_disagreement_score":0.04588807,"threshold_uncertainty_score":0.1535109,"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."}}