Integrative analysis of transcriptome and proteome-wide association study identifies novel genes implicated in tourette's syndrome

Tourette Syndrome (TS) is a neurodevelopmental disorder characterized by involuntary motor and phonic tics. Genome-wide association studies (GWAS) have advanced the genetic understanding of TS by identifying genome-wide loci significantly associated with TS. However, determining the biological mechanisms and pathways of GWAS signals remains challenging. To characterize the effect of genetic variation-mediated gene expression in TS and to understand the biological underpinnings of the disorder, we performed a global and unbiased transcriptome-wide and proteome-wide association study (TWAS and PWAS, respectively) in the largest cohort of TS patient samples of general European ancestry consisting of 13,247 TS cases and 536,217 healthy controls. Transcriptomic imputation was performed using the FUSION tool to predict tissue-specific gene expression based on the latest GWAS summary statistics of TS (N total = 549,464; N effective = 41,204). Precomputed gene expression weights from 49 GTEx tissues (version 8) was used to estimate a gene's association with TS. As transcriptome expression sparingly correlates with proteome expression, we performed PWAS to complement TWAS analysis. The FUSION pipeline was used to perform PWAS analysis based on protein expression weights from Dorso-Lateral Prefrontal Cortex (DLPFC) tissue. Gene associations were considered significant if the TWAS or PWAS p-value was less than 0.05/number of genes per tissue. Additionally, colocalization analysis was computed for genes with a p-value less than 1 × 10E-5 using the COLOC function. Genes with a posterior probability value equal to or greater than 0.75 were considered evidence for the expression quantitative trait loci (eQTL)-GWAS pair influencing both the expression and the GWAS trait in a particular region. TWAS analysis using individual tissue-based prediction matrix of gene expression identified 60 unique genes whose transcript expression was significantly associated with TS. Notably, TWAS identifies additional loci that are not significant at the genome-wide levels. PWAS analysis, based on protein expression from DLPFC tissue, identified two unique genes. Interestingly, both genes identified by PWAS were also implicated using the TWAS method suggesting them to be linked with TS, making them key findings of the study. For these two genes, genetic variant-mediated change in transcript and protein expression is significantly associated with TS etiology. In conclusion, results from our TWAS and PWAS analysis allow us to identify novel genes associated with TS disorder and identify biological pathways that can be validated via biological experimentation to strengthen our analysis.