{"id":"W2112072385","doi":"10.1093/bioinformatics/btv006","title":"Integrating alignment-based and alignment-free sequence similarity measures for biological sequence classification","year":2015,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":43,"is_retracted":false,"has_abstract":true,"ca_institutions":"Ontario Institute for Cancer Research","funders":"Government of Ontario; Ontario Institute for Cancer Research","keywords":"Weighting; Similarity (geometry); Sequence alignment; Sequence (biology); Classifier (UML); Multiple sequence alignment; Computer science; Artificial intelligence; Pattern recognition (psychology); Set (abstract data type); Metagenomics; Protein sequencing; Data mining; Computational biology; Biology; Genetics; Peptide sequence","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":[0.000443833,0.0001666439,0.0001488203,0.00002535455,0.0001156385,0.00004636768,0.0002321361,0.0001367631,7.228963e-7],"category_scores_gemma":[0.0005148692,0.0001347883,0.00005129246,0.00004502321,0.0001559612,0.000003216432,0.0001360292,0.00004699218,0.000001717737],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003529925,"about_ca_system_score_gemma":0.0001128229,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001081329,"about_ca_topic_score_gemma":0.00001573356,"domain_scores_codex":[0.9991094,0.00003105904,0.0003057203,0.0001935536,0.0001360463,0.0002242272],"domain_scores_gemma":[0.9992037,0.00003328661,0.0001454106,0.0003398743,0.0001628048,0.000114874],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.000226748,0.0001141944,0.02190701,0.0001580643,0.0001609684,0.000001127705,0.001068847,0.0002446649,0.9361954,0.005196055,0.009366571,0.02536031],"study_design_scores_gemma":[0.009736438,0.005538036,0.01049973,0.0001682062,0.0002473581,0.00007206303,0.01364711,0.1786659,0.4935911,0.01663537,0.2684072,0.002791515],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9155499,0.0009144553,0.07960206,0.0009176476,0.0002618463,0.0008718151,0.0004358057,0.00002298637,0.001423471],"genre_scores_gemma":[0.9209817,0.0001492643,0.07797702,0.000568091,0.00007714245,0.00006635107,0.0001459974,0.000009595386,0.000024853],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4426044,"threshold_uncertainty_score":0.5496507,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1221202837372235,"score_gpt":0.3031517359470896,"score_spread":0.181031452209866,"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."}}