{"id":"W2564054800","doi":"10.1093/ve/vew031","title":"Impacts and shortcomings of genetic clustering methods for infectious disease outbreaks","year":2016,"lang":"en","type":"article","venue":"Virus Evolution","topic":"HIV Research and Treatment","field":"Immunology and Microbiology","cited_by":93,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University","funders":"","keywords":"Cluster analysis; Context (archaeology); Computer science; Outbreak; Population; Data mining; Data science; Machine learning; Geography; Biology; Medicine; Environmental health; Virology","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0610996,0.001517381,0.001653852,0.003815632,0.002227722,0.00434903,0.004348386,0.003844033,0.001712129],"category_scores_gemma":[0.2549668,0.00084125,0.002198717,0.004764351,0.004156031,0.004867572,0.002836328,0.003794589,0.00113034],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004234711,"about_ca_system_score_gemma":0.004924774,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01935611,"about_ca_topic_score_gemma":0.01347874,"domain_scores_codex":[0.9394533,0.04357644,0.001702463,0.004338748,0.01050097,0.0004279739],"domain_scores_gemma":[0.7654842,0.1880904,0.006945434,0.01585187,0.02245872,0.001169388],"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":[0.0005788031,0.000172663,0.06665614,0.001325221,0.001530382,0.00024486,0.002367648,0.3217233,0.001463256,0.1321686,0.01454521,0.4572239],"study_design_scores_gemma":[0.0001426135,0.0002315385,0.01891751,0.001038601,0.0002762348,0.0006544476,0.001144022,0.7673655,0.003404246,0.1841422,0.02243135,0.0002518007],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06305455,0.01556193,0.8831545,0.02234081,0.001276568,0.0004360445,0.0006101604,0.001185953,0.01237938],"genre_scores_gemma":[0.4683418,0.009042045,0.5135912,0.00297672,0.00120595,0.0004526572,0.0007429693,0.0007567706,0.002889981],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.0610996,"threshold_uncertainty_score":0.3231295,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0177001242066557,"score_gpt":0.3250958457405654,"score_spread":0.3073957215339098,"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."}}