Integrative multi-omics analysis to gain new insights into COVID-19

Abstract

Multidimensional host and viral factors determine the clinical course of COVID-19. While the virology of the disease is well studied, investigating host-related factors, including genome, transcriptome, metabolome, and exposome, can provide valuable insights into the underlying pathophysiology. We conducted integrative omics analyses to explore their intricate interplay in COVID-19. We used data from the UK Biobank (UKB), and employed single-omics, pairwise-omics, and multi-omics models to illustrate the effects of different omics layers. The dataset included COVID-19 phenotypic data as well as genome, imputed-transcriptome, metabolome and exposome data. We examined the main, interaction effects and correlations between omics layers underlying COVID-19. Single-omics analyses showed that the transcriptome (derived from the coronary artery tissue) and exposome captured 3–4% of the variation in COVID-19 susceptibility, while the genome and metabolome contributed 2–2.5% of the phenotypic variation. In the omics-exposome model, where individual omics layers were simultaneously fitted with exposome data, the contributions of genome and metabolome were diminished and considered negligible, whereas the effects of the transcriptome showed minimal change. Through mediation analysis, the findings revealed that exposomic factors mediated about 60% of the genome and metabolome’s effects, while having a relatively minor impact on the transcriptome, mediating only 7% of its effects. In conclusion, our integrative-omics analyses shed light on the contribution of omics layers to the variance of COVID-19.