Integrated analysis of COVID-19 multi-omics data for eQTLs reveals genetic mechanisms underlying disease severity

The global pandemic caused by the SARS-CoV-2 virus provided an unprecedented opportunity to investigate genetic factors influencing the disease severity of the viral infection. Despite a plethora of recent research on both SARS-CoV-2 and COVID-19, few have taken a systems biology approach to address individual-level variation, especially based on non-European populations. We analyzed multi-omics data generated at three timepoints from 184 Korean COVID-19 patients with mild or severe symptoms, composed of whole genome sequencing, blood-based single-cell RNA-sequencing (2.15M cells), 195 cytokine profiles, and human leukocyte antigen (HLA) allele data. Expression quantitative trait loci (eQTLs) were called and their biological implications were evaluated. We identified disease severity interacting eQTLs (n = 404) and disease progression interacting eQTLs (n = 972) for various cell types. We elucidated a complex regulatory mechanism involving HLA genes and their targets, and identified genetic determinants of cytokine levels. Finally, we show how regulation of interaction-eQTLs (ieQTLs) is established by upstream transcription factors (TFs), illustrating complex regulation of the IGFBP7 ieQTL, which is potentially important in conferring differential severity. Our COVID-eQTLs were independently validated against a large-scale single-cell PBMC eQTL dataset (1M-scBloodNL), providing robust cross-dataset replication across cell types. This study illuminates an efficient molecular interrogation framework that can be applied toward understanding infectious disease progression in individuals of different genotypes.