Fig. 1
From: Integrative multi-omics analysis to gain new insights into COVID-19
A diagram of study design for partitioning of COVID-19 phenotypic variance using multi-omics data. The phenotypic variance partitioning analyses were based on multi-omics layers, including genome (G), transcriptome (T), metabolome (M) and exposome (E). Phenotypic variance also captured by omics-exposome interactions, such as genome-exposome (GxE), transcriptome-exposome (TxE), and metabolome-exposome (MxE). Interactions between omics layers, genome-transcriptome (GxT), and transcriptome-metabolome (TxM) were also explored to examine variance components. Furthermore, it is worth noting that due to complex interplays between omics layers, part of the phenotypic variation can be explained by covariance structures, thus revealing omics correlations. Correlation structures were sought between genome & exposome (rG,E), genome & transcriptome (rG,T), transcriptome & exposome (rT,E), transcriptome & metabolome (rT,M), and metabolome & exposome (rM,E). Apart from metabolomics data (only 23520 samples were accessed), other omics data were obtained from 107857 UKB participants. Genome and metabolome interplay was not carried out because the effect of the genome was negligible, based on the effect of metabolome-matched cases on phenotypic variation.
