{"id":"W2169569956","doi":"10.1155/2015/196591","title":"Spaced Seed Data Structures for<i>De Novo</i>Assembly","year":2015,"lang":"en","type":"article","venue":"International Journal of Genomics","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canada's Michael Smith Genome Sciences Centre","funders":"National Human Genome Research Institute; National Institutes of Health; Genome British Columbia; Genome Canada","keywords":"Sequence assembly; Bloom filter; Data structure; k-mer; Genome; Computer science; De Bruijn graph; De Bruijn sequence; Hybrid genome assembly; Scalability; Sequence (biology); Metagenomics; De novo transcriptome assembly; Reference genome; Algorithm; Computational biology; Biology; Theoretical computer science; Graph; Genetics; Transcriptome; Mathematics; Gene; Database; Programming language; Combinatorics","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":[0.000336125,0.00009110528,0.000118196,0.00004457465,0.00002575007,0.000042482,0.000798384,0.00005977616,0.000001428121],"category_scores_gemma":[0.0002198264,0.00008348691,0.00006986157,0.0000167207,0.00003253219,0.000002419157,0.0002651897,0.00005561738,9.759817e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004489615,"about_ca_system_score_gemma":0.000387712,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000008528662,"about_ca_topic_score_gemma":0.00002436074,"domain_scores_codex":[0.9993309,0.00001720397,0.0002477175,0.0001351886,0.0001457575,0.0001232569],"domain_scores_gemma":[0.9988645,0.00001666218,0.0002128439,0.0002117149,0.0006042715,0.00008999115],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0007110099,0.00004937063,0.006982145,0.000004063305,0.0006915342,0.00001431047,0.0002629743,0.002812258,0.9743791,0.000539689,0.01078656,0.002767031],"study_design_scores_gemma":[0.00404427,0.0007138202,0.01113001,0.00001223862,0.0001021869,0.000853853,0.0006240177,0.0007706769,0.1229962,0.006599187,0.8518044,0.0003491752],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9873055,0.001328084,0.008824541,0.0008817796,0.001177548,0.00006424999,0.000168857,9.457448e-7,0.000248516],"genre_scores_gemma":[0.9816253,0.0003189426,0.01584087,0.0003903933,0.001603135,0.000001131768,0.00007084312,0.00001632082,0.000133022],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8513829,"threshold_uncertainty_score":0.3404497,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04572948246611166,"score_gpt":0.3164276191650766,"score_spread":0.270698136698965,"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."}}