{"id":"W2148425737","doi":"10.1093/bioinformatics/btu558","title":"BioBloom tools: fast, accurate and memory-efficient host species sequence screening using bloom filters","year":2014,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":134,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canada's Michael Smith Genome Sciences Centre","funders":"Genome British Columbia; Genome Canada","keywords":"Bloom filter; False positive paradox; Computer science; Set (abstract data type); Sequence (biology); Filter (signal processing); Host (biology); Software; False positives and false negatives; Data set; Data structure; Data mining; Algorithm; Artificial intelligence; Programming language; Biology","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":[0.0002134751,0.0001962645,0.0001751714,0.0000469205,0.0002001609,0.0001082655,0.0001644573,0.00008813606,0.000004916603],"category_scores_gemma":[0.00009738307,0.0001753139,0.0000583599,0.00007255666,0.0001586716,0.000003592082,0.0002575504,0.00006055745,0.00000444616],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000009244109,"about_ca_system_score_gemma":0.00002928078,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001139065,"about_ca_topic_score_gemma":0.000005800895,"domain_scores_codex":[0.9990861,0.00002368201,0.0002985023,0.0001823455,0.0001219897,0.0002873792],"domain_scores_gemma":[0.9993823,0.00002576186,0.0001474146,0.0002862744,0.00006976294,0.00008855355],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00005639967,0.00004977878,0.002340376,0.0001804719,0.0002070384,0.000002292583,0.0016414,0.01291802,0.9417077,0.0002709495,0.0007666497,0.03985888],"study_design_scores_gemma":[0.002620594,0.0009001035,0.01836877,0.0001910601,0.0002200904,0.0001657581,0.006077466,0.5141926,0.3545921,0.00005530751,0.100703,0.001913126],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9623002,0.0004363554,0.03512798,0.00007746464,0.0001620509,0.0001724756,0.00008286147,0.000008465372,0.001632128],"genre_scores_gemma":[0.9606536,0.0002467461,0.03835176,0.0003522268,0.0001757756,0.000004637796,0.00003370461,0.00001702264,0.0001644987],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5871156,"threshold_uncertainty_score":0.7149092,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04576000212109506,"score_gpt":0.2551211840184869,"score_spread":0.2093611818973918,"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."}}