{"id":"W2001390039","doi":"10.1089/cmb.2006.0136","title":"Efficient Algorithms for Counting and Reporting Segregating Sites in Genomic Sequences","year":2007,"lang":"en","type":"article","venue":"Journal of Computational Biology","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"McMaster University","funders":"","keywords":"Set (abstract data type); Sequence (biology); Algorithm; Computer science; Sample (material); Sublinear function; Data set; Biology; Computational biology; Mathematics; Genetics; Artificial intelligence; Combinatorics","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001367236,0.00006888148,0.000163814,0.00006873024,0.00005913317,0.000008796893,0.00005393017,0.0000543924,5.240744e-7],"category_scores_gemma":[0.0003641403,0.00005886031,0.00004965627,0.00003882746,0.00006032297,4.691177e-7,0.00003804162,0.00005140352,9.354849e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001347521,"about_ca_system_score_gemma":0.00006092578,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000004855506,"about_ca_topic_score_gemma":0.00001161888,"domain_scores_codex":[0.9989883,0.00001975613,0.0006897958,0.000117639,0.00004520823,0.0001392726],"domain_scores_gemma":[0.9987478,0.000154795,0.0008460992,0.00002962204,0.0001938031,0.00002783873],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"observational","study_design_scores_codex":[0.00007775716,0.00002530643,0.2063289,0.00001664381,0.00007003436,0.000006789284,0.0001926376,0.06990448,0.7134128,0.0002914742,0.00001664551,0.009656533],"study_design_scores_gemma":[0.002681247,0.00161963,0.9070429,0.00006886837,0.0000472859,0.001030488,0.001230712,0.03830402,0.02724328,0.01745503,0.002797305,0.0004792351],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9782844,0.001933606,0.0194543,0.00009722614,0.0001165102,0.00006917409,0.000004907344,5.589728e-7,0.00003930174],"genre_scores_gemma":[0.9529151,0.0000252474,0.04670719,0.00008832355,0.0002469,0.000001036085,0.000007877168,0.000004853154,0.000003495359],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.700714,"threshold_uncertainty_score":0.2400254,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0333006308644674,"score_gpt":0.3234489920134218,"score_spread":0.2901483611489544,"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."}}