{"id":"W1996640618","doi":"10.1016/j.cmpb.2005.11.007","title":"n-Gram-based classification and unsupervised hierarchical clustering of genome sequences","year":2006,"lang":"en","type":"article","venue":"Computer Methods and Programs in Biomedicine","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":98,"is_retracted":false,"has_abstract":false,"ca_institutions":"Dalhousie University","funders":"","keywords":"Substring; Hierarchical clustering; Cluster analysis; Preprocessor; Computer science; Euclidean distance; Distance measures; Genome; Pattern recognition (psychology); Gram; Artificial intelligence; n-gram; Data mining; Data structure; Biology; Genetics; Gene","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":[0.0004744113,0.0001128324,0.0001957013,0.00007268615,0.00003807335,0.00001303234,0.00007718976,0.00007816257,7.660797e-7],"category_scores_gemma":[0.000007643961,0.00009125411,0.00002509587,0.0001179725,0.000335771,6.677955e-7,0.00008849158,0.0000497057,4.389536e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000004014315,"about_ca_system_score_gemma":0.00001634628,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006733433,"about_ca_topic_score_gemma":0.00002014252,"domain_scores_codex":[0.9991693,0.0001114629,0.0002486499,0.0002693973,0.00005739692,0.0001438054],"domain_scores_gemma":[0.9996888,0.00003560799,0.00005947603,0.0001377926,0.00003586126,0.00004247681],"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":[0.0000395808,0.00006136278,0.02980858,0.00009371493,0.00002002329,0.000001473165,0.00008421015,0.00005086621,0.5683063,0.0001215068,0.000006150171,0.4014062],"study_design_scores_gemma":[0.003411807,0.002884372,0.8742229,0.0001468713,0.00005904418,0.00003944651,0.0001695858,0.07468155,0.01132826,0.003362401,0.02915417,0.0005396305],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8053089,0.004323595,0.1897164,0.0003157232,0.0000863039,0.0001920745,0.000003713877,0.000003450541,0.00004982472],"genre_scores_gemma":[0.6643916,0.000234882,0.3351105,0.0000682048,0.000132845,0.00001235126,0.00003878056,0.000005865583,0.000004899974],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8444143,"threshold_uncertainty_score":0.3721235,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04474609928687936,"score_gpt":0.3246002451323166,"score_spread":0.2798541458454373,"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."}}