{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005754065,0.001287743,0.001374677,0.005126124,0.0006586735,0.002253037,0.002422217,0.001923267,0.001636066],"category_scores_gemma":[0.01643543,0.000346949,0.001346005,0.004337546,0.001166489,0.004530246,0.002031194,0.002261745,0.001890817],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001514706,"about_ca_system_score_gemma":0.001305635,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002693151,"about_ca_topic_score_gemma":0.002818364,"domain_scores_codex":[0.9957018,0.001104156,0.0004097823,0.001092353,0.001514691,0.0001771538],"domain_scores_gemma":[0.9914671,0.004236529,0.001070246,0.001047279,0.001770649,0.0004082395],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0008716943,0.0007983622,0.0535987,0.001211928,0.0006529809,0.0002546312,0.0004680912,0.2002492,0.03440545,0.0211922,0.01151118,0.6747856],"study_design_scores_gemma":[0.00002674693,0.0002034377,0.004867818,0.00007852886,0.00007241886,0.0001740781,0.0000685027,0.9459535,0.008733617,0.03475815,0.005010593,0.00005262011],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.06315179,0.001941614,0.9268516,0.0007228056,0.000186136,0.0002862728,0.001495571,0.002875872,0.002488391],"genre_scores_gemma":[0.4849012,0.0009998638,0.5063729,0.0004346007,0.0003590842,0.0004974648,0.004338598,0.0003825745,0.001713809],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005754065,"threshold_uncertainty_score":0.03043079,"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."}}