Rapid screening and identification of viral pathogens in metagenomic data

• Published in: BMC medical genomics 2021-12, Vol.14 (Suppl 6), p.289-11, Article 289
• Main Authors: Song, Shiyang, Ma, Liangxiao, Xu, Xintian, Shi, Han, Li, Xuan, Liu, Yuanhua, Hao, Pei
• Format: Article
• Language: English
• Subjects: Algorithms; Animals; Automation; China; Computational Biology - methods; Control; Coronavirus Infections - genetics; Coronaviruses; COVID-19; COVID-19 - diagnosis; COVID-19 Testing - methods; Disease transmission; Epidemic; Epidemics; Genome, Viral; Genomes; Genomics; High-Throughput Nucleotide Sequencing; Humans; Identification; Identification and classification; Mass Screening - methods; Metagenomics; Metagenomics data; Mutation; Next-generation sequencing; Pandemics; Pangolins; Pathogen screening; Pathogenic microorganisms; Pathogens; Pneumonia; Reference Values; SARS-CoV-2; SARS-CoV-2 - genetics; Severe acute respiratory syndrome coronavirus 2; Software; Statistical analysis; Taxonomy; Transcriptome; Viral genome assembly; Viruses; Workflow
• ISSN: 1755-8794; 1755-8794
• DOI: 10.1186/s12920-021-01138-z

Virus screening and viral genome reconstruction are urgent and crucial for the rapid identification of viral pathogens, i.e., tracing the source and understanding the pathogenesis when a viral outbreak occurs. Next-generation sequencing (NGS) provides an efficient and unbiased way to identify viral pathogens in host-associated and environmental samples without prior knowledge. Despite the availability of software, data analysis still requires human operations. A mature pipeline is urgently needed when thousands of viral pathogen and viral genome reconstruction samples need to be rapidly identified. In this paper, we present a rapid and accurate workflow to screen metagenomics sequencing data for viral pathogens and other compositions, as well as enable a reference-based assembler to reconstruct viral genomes. Moreover, we tested our workflow on several metagenomics datasets, including a SARS-CoV-2 patient sample with NGS data, pangolins tissues with NGS data, Middle East Respiratory Syndrome (MERS)-infected cells with NGS data, etc. Our workflow demonstrated high accuracy and efficiency when identifying target viruses from large scale NGS metagenomics data. Our workflow was flexible when working with a broad range of NGS datasets from small (kb) to large (100 Gb). This took from a few minutes to a few hours to complete each task. At the same time, our workflow automatically generates reports that incorporate visualized feedback (e.g., metagenomics data quality statistics, host and viral sequence compositions, details about each of the identified viral pathogens and their coverages, and reassembled viral pathogen sequences based on their closest references). Overall, our system enabled the rapid screening and identification of viral pathogens from metagenomics data, providing an important piece to support viral pathogen research during a pandemic. The visualized report contains information from raw sequence quality to a reconstructed viral sequence, which allows non-professional people to screen their samples for viruses by themselves (Additional file 1).