Crowdsourcing Coronavirus Data: Pt4 Data Visualization

"Crowdsourcing Coronavirus Data" by Mehdi Ahmadi1, James Flynn2, Joe Deleney2, Enrique Torchia1 and Michael Edwards3. 1. Department of Dermatology, University of Colorado Anschutz Medical Campus, Aurora,CO. 2. Illumina Corporation, City. 3. Bioinfo Solutions LLC, Parker, CO Background To date, little is known about the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that is at the center of the current global epidemic. (Lai et al., 2020; Shereen et al., 2020). SARS-CoV-2 belongs to a new strain of human viruses that emerged from Wuhan providence in China in December 2019. There are currently no animal studies addressing the host biology of SARS-CoV-2 infection. However, the virus does shares 79% homology to previously identified human infecting coronavirus (e.g., SARS CoV1) including binding to ACE2 receptor on susceptible cells (Lu et al., 2020; Wu et al., 2020). These observations provide the rationale for the mining of previous studies and their public available datasets on coronavirus infection to understand the current viral outbreak. Data Collection Illumina’s Correlation Engine software was used to identify and analyze publicly accessible data from the Gene Expression omnibus depository (GEO)(www.ncbi.nlm.nih.gov/geo/). Relevant and curated biosets were identified by the correlation engine app using the search term “Coronavirus infection”. We focused on the host gene expression responses during viral infection because of the abundance of available published data and the potential to identify possible therapeutic interventions. We identified 39 biosets (comparisons) from 14 different studies that measured changes in global gene expression in mouse and monkey lungs collected at 1, 4 and 7 days post-infection relative to mock infection controls. A list of biosets and associated information used for this analysis can be found and downloaded as excel or text files from GitHub at the following link: https://github.com/BioinfoSolutions/C.... Data Processing A meta-dataset was created with the correlation engine app using a ranked-based gene score for each bioset and time point. These data were then downloaded from the correlation engine software and further processed by a python-based algorithm that allowed the conversion of the ranked Illumina score data into positive and negative values based on the fold change. See Cheema et al. (Cheema et al., 2020) and Kupershmidt et al. (Kupershmidt et al., 2010) for further details on data processing and statistical methods used by correlation engine. References Cheema SK, Isesele PO, Marchando S, et al. (2020) The suppression of very long chain fatty acids is associated with skin carcinogenesis. Journal of Investigative Dermatology. In press. Kupershmidt I, Su QJ, Grewal A, et al. (2010) Ontology-based meta-analysis of global collections of high-throughput public data. PLoS One 5. Lai C-C, Liu YH, Wang C-Y, et al. (2020) Asymptomatic carrier state, acute respiratory disease, and pneumonia due to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2): Facts and myths. Journal of Microbiology, Immunology and Infection https://doi.org/10.1016/j.jmii.2020.0.... Lu R, Zhao X, Li J, et al. (2020) Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding. The Lancet 395:565-74. Shereen MA, Khan S, Kazmi A, et al. (2020) COVID-19 infection: Origin, transmission, and characteristics of human coronaviruses. Journal of Advanced Research 24:91-8. Wu A, Peng Y, Huang B, et al. (2020) Genome Composition and Divergence of the Novel Coronavirus (2019-nCoV) Originating in China. Cell Host & Microbe 27:325-8.