Genome Detective Coronavirus Typing Tool for rapid identification and characterization of novel coronavirus genomes

Sara Cleemput; Wim Dumon; Vagner Fonseca; Wasim Abdool Karim; Marta Giovanetti; Luiz Carlos Alcantara; Koen Deforche; Tulio de Oliveira

This coronavirus outbreak has been unprecedented; so too is the way that the scientific community has responded to it. They have openly and rapidly shared genomic and clinical data as never seen before allowing research results to be released almost instantaneously. This has helped the understanding of the transmission dynamics, the development of rapid diagnostic and has informed public health response. Here, we present a new contribution that can speed up this communal effort. The Genome Detective Coronavirus Typing Tool is a free-of-charge web-based bioinformatics pipeline that can accurately and quickly identify, assemble and classify coronaviruses genomes. The tool also identifies changes at nucleotides, coding regions and proteins using a novel dynamic aligner to allow tracking new viral mutations (Figure 1 ). To this reference dataset, we added 47 whole genomes of the current Coronavirus 2019 (2019-nCoV) outbreak that originated in Wuhan, China, in December 2019. The 2019-nCoV sequences were downloaded from the GISAID database (https://www.gisaid.org) together with annotation of its original location, collection date and originating and submitting laboratory. The 2019-nCoV data generators are properly acknowledged in the acknowledgements section of this paper and detailed information is provided in Supplementary Table 1. The 431 reference WGS were aligned with MUSCLE (Edgar 2004) . The alignment was manually edited until a codon alignment was attained in all coding sequences (CDS). A Maximum likelihood phylogenetic tree, 1000 bootstrap replicates was constructed in PhyML (Guidon & Gascuel 2003; Lemoine et al., 2018) and a Bayesian tree using MrBayes (Ronquist & Huelsenbeck 2003) were constructed. The trees were visualized in Figtree (Rambaut 2018). We selected 25 reference sequences that represent the diversity of each well-defined phylogenetic cluster (with bootstrap support of 100% and posterior probability of 1). We identified five well supported phylogenetic clusters with more than two sequences of SARSr-CoV and used them to set up our automated phylogenetic classification tool. Cluster 1 included SARS strains from the 2002 and 2003 Asian outbreaks. In our tool, we named this cluster SARS-CoV Outbreak 2000s but may rename it as SARS-A if a new proposed naming system for SARSr-Cov is adopted in the near future (Rambaut 2020). Cluster 2 (provisionally named as SARS related CoV) includes 7 sequences from bats which did not cause large human outbreaks. Cluster 3 (named as Bat SARS-CoV HKU3) includes three WGS sampled from Rhinolophus sinicus (i.e. Chinese rufous horseshoe bats). Cluster 4 (Bat SARS-CoV ZXC21/ZC45) includes two SARSr-CoV sampled from Rhinolophus sinicus bats in Zhoushan, China. Cluster 5 (provisionally named as Wuhan 2019-nCoV, which may be renamed as SARS-B) includes one public sequence from the outbreak in Wuhan, China. We identified this cluster with many sequences from GISAID but kept only this one as this is the only GenBank sequence, accession number MN908947, which was kindly shared by Prof. Yong-Zhen Zhang and colleagues in the virological.org website. Detailed information about the phylogenetic reference datasets are available in Supplementary Table 2. The phylogenetic reference dataset was used to create an automated Coronavirus Typing Tool using the Genome Detective framework (Vilsker et al., 2019 . To determine the accuracy of this tool, each of the 431 test WGS was considered for evaluation (i.e. 384