{"id":"W4390366463","doi":"10.1101/2023.12.28.573569","title":"Eukfinder: a pipeline to retrieve microbial eukaryote genomes from metagenomic sequencing data","year":2023,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Parasitic Infections and Diagnostics","field":"Immunology and Microbiology","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"Alliance de recherche numérique du Canada; University of Toronto; Innovation, Science and Economic Development Canada","keywords":"Metagenomics; Biology; Genome; Computational biology; Blastocystis; Microbiome; Metaproteomics; Shotgun sequencing; Eukaryote; Sequence assembly; Workflow; Bacterial genome size; Evolutionary biology; Genetics; Gene; Ecology; Computer science; Transcriptome; Database","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":[0.00312411,0.002908502,0.00132498,0.004606313,0.001127129,0.001990803,0.003039635,0.001396597,0.005481089],"category_scores_gemma":[0.004757592,0.001752774,0.002083953,0.002183465,0.0005444083,0.002520995,0.002711713,0.001482288,0.003997331],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007142346,"about_ca_system_score_gemma":0.001733852,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00186729,"about_ca_topic_score_gemma":0.002266729,"domain_scores_codex":[0.9988432,0.0001957698,0.0001352874,0.0004575907,0.0002646997,0.0001034283],"domain_scores_gemma":[0.9987918,0.0004410522,0.0002478966,0.0001994983,0.0001912078,0.0001285834],"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":[0.004771978,0.0006344148,0.01958675,0.007384735,0.001704768,0.001890946,0.001537726,0.03232074,0.4410374,0.007346479,0.1144383,0.3673458],"study_design_scores_gemma":[0.0009890122,0.0006436042,0.02128496,0.0006346218,0.0004164105,0.001374387,0.0005243405,0.500689,0.3259009,0.01808787,0.128765,0.000690005],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"software","genre_scores_codex":[0.04237177,0.001534262,0.6572315,0.0003766812,0.000165707,0.0007024863,0.05312125,0.2426894,0.001806937],"genre_scores_gemma":[0.08858085,0.0006303843,0.8160189,0.0003604673,0.00005048842,0.00121165,0.08252324,0.00874598,0.001877991],"genre_candidate":"software","genre_consensus":null,"teacher_disagreement_score":0.005481089,"threshold_uncertainty_score":0.01833606,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04652650975856265,"score_gpt":0.2584994644308347,"score_spread":0.2119729546722721,"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."}}