{"id":"W2119894516","doi":"10.1186/1471-2105-8-286","title":"CLUSS: Clustering of protein sequences based on a new similarity measure","year":2007,"lang":"en","type":"article","venue":"BMC Bioinformatics","topic":"Machine Learning in Bioinformatics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":59,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Sherbrooke","funders":"Université de Sherbrooke","keywords":"Cluster analysis; Similarity (geometry); Phylogenetic tree; Biological data; Similarity measure; Computer science; Matching (statistics); Protein sequencing; Measure (data warehouse); Data mining; Protein family; Computational biology; Artificial intelligence; Bioinformatics; Biology; Genetics; Peptide sequence; Mathematics; Gene","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":[0.001748539,0.001657423,0.001850312,0.008165902,0.001503199,0.001793709,0.002308606,0.00148215,0.002921163],"category_scores_gemma":[0.006044791,0.0005888866,0.00194801,0.006287765,0.001552612,0.002928126,0.002628414,0.001342028,0.002231402],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001351156,"about_ca_system_score_gemma":0.001668206,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001734367,"about_ca_topic_score_gemma":0.001598451,"domain_scores_codex":[0.9966424,0.0005742472,0.0003922996,0.0006878562,0.001540413,0.000162834],"domain_scores_gemma":[0.9973754,0.0006097847,0.0005150596,0.0004382458,0.0008488234,0.0002127727],"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":[0.001505477,0.0005070785,0.01904672,0.002206464,0.001221548,0.0006455495,0.00115858,0.123676,0.07126323,0.03506672,0.03547876,0.7082238],"study_design_scores_gemma":[0.0002368066,0.0004528372,0.005945177,0.000144506,0.000181709,0.001004517,0.0003919245,0.8748063,0.03680698,0.03936663,0.04050222,0.0001604448],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03803408,0.001095186,0.9485748,0.0003310688,0.0002763906,0.0004702397,0.00175949,0.007433652,0.002025103],"genre_scores_gemma":[0.1214093,0.0004935263,0.8698924,0.0001332208,0.0001674402,0.0006380202,0.005070807,0.0007482748,0.001447046],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.008165902,"threshold_uncertainty_score":0.009803414,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01812383730328375,"score_gpt":0.2674233571612362,"score_spread":0.2492995198579524,"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."}}