{"id":"W4389525196","doi":"10.1093/molbev/msad270","title":"A Fast, Reproducible, High-throughput Variant Calling Workflow for Population Genomics","year":2023,"lang":"en","type":"article","venue":"Molecular Biology and Evolution","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":70,"is_retracted":false,"has_abstract":true,"ca_institutions":"Mount Royal University","funders":"Division of Biological Infrastructure; FAS Division of Science, Harvard University; Harvard University; National Science Foundation","keywords":"Workflow; Biology; Genomics; Reuse; Pipeline (software); Population; Flexibility (engineering); Population genomics; Cloud computing; Computer science; Genome; Computational biology; Data mining; Data science; Database; Genetics; Gene; Ecology","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":[0.008345895,0.001247368,0.001295923,0.002323549,0.002130912,0.002465639,0.002660092,0.001402943,0.006103808],"category_scores_gemma":[0.01181266,0.001406159,0.002731017,0.00253482,0.0009170207,0.001570141,0.002872869,0.003899009,0.00625007],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001080444,"about_ca_system_score_gemma":0.003912517,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003472946,"about_ca_topic_score_gemma":0.007192599,"domain_scores_codex":[0.9957162,0.0007880646,0.0004605595,0.001337325,0.00144269,0.0002552655],"domain_scores_gemma":[0.995545,0.001259054,0.0004564485,0.001446351,0.001030753,0.0002623977],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001121081,0.0004141716,0.01403144,0.001571596,0.001096129,0.0006920443,0.001244833,0.02095727,0.4706454,0.01820025,0.06632239,0.4037035],"study_design_scores_gemma":[0.0005642608,0.0006770155,0.04026423,0.0005208595,0.0005495678,0.003049382,0.0004293001,0.1814943,0.4302517,0.07378215,0.2672929,0.0011244],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01351956,0.0003585727,0.9493209,0.0002151758,0.0001953412,0.0004831012,0.007312578,0.02694975,0.001645171],"genre_scores_gemma":[0.03459769,0.0003068497,0.9415359,0.000228799,0.00005527186,0.0008730529,0.01623051,0.003820094,0.002351669],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.008345895,"threshold_uncertainty_score":0.04413778,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01095052351500947,"score_gpt":0.2587298556143819,"score_spread":0.2477793320993724,"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."}}