{"id":"W4409333289","doi":"10.1128/mbio.00699-25","title":"Eukfinder: a pipeline to retrieve microbial eukaryote genome sequences from metagenomic data","year":2025,"lang":"en","type":"article","venue":"mBio","topic":"Microbial Community Ecology and Physiology","field":"Environmental Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"Canadian Institutes of Health Research","keywords":"Metagenomics; Genome; Biology; Computational biology; Microbiome; Human Microbiome Project; Bacterial genome size; Sequence assembly; Shotgun sequencing; Genomics; Genetics; Gene; Transcriptome","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":["insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.0003405274,0.0001484921,0.0002355612,0.00004422828,0.0002725233,0.0000260695,0.001562366,0.000135142,0.009784376],"category_scores_gemma":[0.00007339924,0.0001461129,0.00003917021,0.0003018676,0.0002115331,0.0001197761,0.002167816,0.0002684411,0.002994212],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001185562,"about_ca_system_score_gemma":0.00004070795,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004762043,"about_ca_topic_score_gemma":0.009078615,"domain_scores_codex":[0.9987395,0.0002085766,0.000234234,0.0004718488,0.00005513213,0.0002906803],"domain_scores_gemma":[0.99866,0.0001222853,0.0000538391,0.001087143,0.000006177999,0.00007061186],"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":[0.00009961098,0.00006956309,0.002456289,0.00000328277,0.00004495946,0.000007387275,0.000348567,0.0001497026,0.8661537,0.00004534177,0.1295062,0.001115399],"study_design_scores_gemma":[0.0006494509,0.0001145534,0.1003851,0.00001304257,0.00009513286,0.000005012704,0.00009656911,0.0001668246,0.0207742,0.005186609,0.8720499,0.0004636503],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9876202,0.0001382995,0.001128553,0.003248492,0.0004094554,0.0002600827,0.0005397763,0.00004534233,0.006609795],"genre_scores_gemma":[0.9825805,0.00003590624,0.002779616,0.009244126,0.0000776807,0.000004706885,0.0009128667,0.00000984038,0.004354732],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8453795,"threshold_uncertainty_score":0.9977821,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0252722375234503,"score_gpt":0.2667538463827485,"score_spread":0.2414816088592981,"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."}}