Benchmark of thirteen bioinformatic pipelines for metagenomic virus diagnostics using datasets from clinical samples

• DOI: 10.1016/j.jcv.2021.104908
• Publication Date: 2021-07-08T17:52:41Z
• Authors: Jutte J.C. de Vries; Julianne R. Brown; Nicole Fischer; Igor A. Sidorov; Sofia Morfopoulou; Jiabin Huang; Bas B. Oude Munnink; Arzu Sayiner; Alihan Bulgurcu; Christophe Rodriguez; Guillaume Gricourt; Els Keyaerts; Leen Beller; Claudia Bachofen; Jakub Kubacki; Samuel Cordey; Florian Laubscher; Dennis Schmitz; Martin Beer; Dirk Hoeper; Michael Huber; Verena Kufner; Maryam Zaheri; Aitana Lebrand; Anna Papa; Sander van Boheemen; Aloys C.M. Kroes; Judith Breuer; F. Xavier Lopez-L...; Eric C.J. Claas
• Keywords: Benchmark; 03 medical and health sciences; SDG 3 - Good Health and Well-being; info:eu-repo/classification/ddc/616; Virology; Viral metagenomics; Humans; ddc:616; 0303 health sciences; Science & Technology; ENCEPHALITIS; Bioinformatic pipelines; Computational Biology; High-Throughput Nucleotide Sequencing; 2725 Infectious Diseases; 3. Good health; ALIGNMENT; Benchmarking; Infectious Diseases; Viruses; 2406 Virology; 570 Life sciences; biology; Metagenomics; Life Sciences & Biomedicine; 10244 Institute of Virology

Metagenomic sequencing is increasingly being used in clinical settings for difficult to diagnose cases. The performance of viral metagenomic protocols relies a large extent on the bioinformatic analysis. In this study, European Society Clinical Virology (ESCV) Network NGS (ENNGS) initiated benchmark pipelines currently virological laboratories.Metagenomic datasets from 13 samples patients with encephalitis or respiratory infections characterized by PCR were selected. analyzed different diagnostic laboratories participating ENNGS members. and classification tools were: Centrifuge, DAMIAN, DIAMOND, DNASTAR, FEVIR, Genome Detective, Jovian, MetaMIC, MetaMix, One Codex, RIEMS, VirMet, Taxonomer. Performance, characteristics, use, user-friendliness these analyzed.Overall, pathogens high loads detected all evaluated pipelines. contrast, lower abundance mixed only 3/13 pipelines, namely MetaMix. Overall sensitivity ranged 80% (10/13) 100% (13/13 datasets). positive predictive value 71-100%. majority classified sequences based nucleotide similarity (8/13), minority amino acid similarity, 6 assembled de novo. No clear differences that correlated approaches. Read counts target viruses varied between over range 2-3 log, indicating limit detection.A wide variety laboratories. Detection low abundant remains challenge, implicating need standardization validation analysis use. Future studies should address selective effects due choice reference databases.