Rapid metagenomic characterization of a case of imported COVID-19 in Cambodia

Jessica E. Manning; Jennifer A. Bohl; Sreyngim Lay; Sophana Chea; Ly Sovann; Yi Sengdoeurn; Seng Heng; Chan Vuthy; Katrina Kalantar; Vida Ahyong; Michelle Tan; Jonathan Sheu; Cristina M. Tato; Joseph L. DeRisi; Laurence Baril; Veasna Duong; Philippe Dussart; Erik A. Karlsson

By February 1, the SARS-CoV-2 genome was correctly identified in the index COVID-19 patient sample using mNGS and globally accessible bioinformatics. An mNGS library was prepared, multiplexed, and sequenced. Notably, the sample generated 4,773,430 reads with 1,903,800 of those being non-host reads. The resulting non-host fastq files were processed using the IDseq platform which utilizes open-source components. When processed on IDseq pipeline version 3.18 with a February 2020 NCBI index, which included 54 SARS-CoV-2 sequences deposited by the international research community, sequences from the index patient aligned to the SARS-CoV-2 reference genome ( Figure 1A , 582 reads, at 100% identity to the reference sequence; NCBI accession MN985325.1). Further analysis resulted in 33.2% coverage of the SARS-CoV-2 reference at an average depth of 2.2x ( Figure 1B ). Assembly and alignment on Geneious resulted in 50.5% coverage at 2.6x of the SARS-CoV-2 genome (NCBI accession MN908947.1). Re-sequencing the index case to a total sequencing depth of 8,786,916 single reads resulted in a greater number of SARS-CoV-2 alignments (819 reads) but had no impact on overall coverage -32.3% coverage breadth at an average depth of 2.7x.