{"id":"W3123592819","doi":"10.1101/2021.01.24.427982","title":"Profile hidden Markov model sequence analysis can help remove putative pseudogenes from DNA barcoding and metabarcoding datasets","year":2021,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Environmental DNA in Biodiversity Studies","field":"Environmental Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph","funders":"Government of Canada; Ontario Genomics; Genome Canada","keywords":"Pseudogene; DNA barcoding; Barcode; Biology; Gene; Amplicon; Computational biology; Genetics; Sequence analysis; Genome; Computer science; Evolutionary biology; Polymerase chain reaction","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":[0.004427554,0.001096254,0.0008763106,0.001458312,0.0008402058,0.001428157,0.001087204,0.001364788,0.001989453],"category_scores_gemma":[0.01468278,0.0005855851,0.002246387,0.001277481,0.0004199859,0.001363953,0.0008874675,0.002120382,0.00122648],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001147078,"about_ca_system_score_gemma":0.002018976,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005126853,"about_ca_topic_score_gemma":0.01156689,"domain_scores_codex":[0.9987748,0.0005001308,0.0001167247,0.0003594805,0.0001764721,0.00007240313],"domain_scores_gemma":[0.9918953,0.006110804,0.0006023666,0.0005384579,0.000681939,0.0001712642],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002916059,0.000908745,0.1303506,0.002593455,0.002304519,0.001013992,0.001167323,0.5243018,0.1021926,0.008928918,0.01588838,0.2074337],"study_design_scores_gemma":[0.00005045743,0.00009073767,0.005436849,0.00004882622,0.0001132363,0.00008605975,0.00006419727,0.9712071,0.01423467,0.006493019,0.002129079,0.00004571567],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.2154389,0.0006014896,0.7531087,0.0007513266,0.00012,0.000249675,0.007920363,0.02071246,0.001097048],"genre_scores_gemma":[0.4140522,0.0002799504,0.5643162,0.0004510877,0.0000493208,0.0004658237,0.01804516,0.001637596,0.0007026367],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005126853,"threshold_uncertainty_score":0.02341539,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02424707748857954,"score_gpt":0.2246199031337322,"score_spread":0.2003728256451526,"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."}}