yu-1011/multi-omics-twas

Integrating genetic regulation of multi-omics levels facilitates elucidating schizophrenia pathogenesis

Multi-Omics-TWAS

Integrating genetic regulation of multi-omics levels elucidates schizophrenia pathogenesis

Workflow

Main results

Genetic regulation are mostly shared from mRNA to protein level

• QTL effect sizes are highly correlated among eQTL, rQTL, and pQTL
  • QTL effect sizes decrease from mRNA to protein level
• QTL P values are highly preserved at mRNA, Ribo-Seq, and protein levels (Pi1)

Genetic regulation signals at different omics levels implicate schizophrenia pathogenesis

• QTL loci at different omics levels contribute to schizophrenia heritability (LDSC)
• Integrating QTL of multi-omics level mediated the most heritability for schizophrenia (MESC)

Running PrediXcan of different omics levels identify schizophrenia risk genes

• Correlation of PrediXcan r2s at different omics levels (Multi-SNP QTL)
• Compare r2 in PrediXcan original and conditional models
• "TWAS" analysis of different omics levels identified unique risk genes, most of the risk genes are exclusively in original models (comparing with conditional models)

Integrating genetic regulation of multi-omics levels identifies schizophrenia risk genes

• meta analysis of TWAS at different omics levels
• colocalization of TWAS signals at different omics levels
• Joint-omics TWAS using MR-JTI

Functional support and possible mechanism behind the schizophrenia risk genes

• HiC, enhancer, etc.
• drug target, known risk genes, MGI, OMIM, GWAS findings, DEGs, pathways, etc.