Identifying the susceptibility genes and variants of neuro-psychiatric diseases will not only contribute to our understanding of these diseases, but also point to potential therapeutic targets. Genome-wide association studies (GWAS) are commonly used to study complex diseases, and have been highly successful in a range of disorder, for instance, more than 100 loci have been associated with the risk of Schizophrenia through GWAS. Nevertheless, in most cases, we do not know the biological mechanisms underlying disease associated loci, because the causal variants and genes are obscured by linkage disequilibrium (LD) and by the difficulty of interpreting functional effects of most genetic variants. The goal of this project is to develop novel statistical methods for integrative analysis of genetic data of neuropsychiatric diseases to better understand the underlying genes and biological processes. (1) We will develop a method to integrate expression QTL (eQTL) data with GWAS. Our method extends the popular Transcriptome-Wise Association Studies (TWAS).
TWAS aims to discover risk genes, by effectively assessing the correlation of eQTLs of a gene with the phenotype of interest. TWAS has many advantages over standard single variant-based analysis, e.g. it reduces multiple testing burden and provides biological contexts of associations. However, current TWAS methods are susceptible to false positive findings. We will develop a rigorous statistical framework to control false discoveries by accounting for pleiotropic effects of variants. (2) Fine-mapping is the statistical approach to identifying causal variants in disease-associated loci. Current fine- mapping methods, however, are often not able to narrow down specific causal variants. Our approach is based on the observation that allelic heterogeneity (AH), i.e. many variants disrupting the same gene, is common. So we can leverage AH to identify risk genes, borrowing the statistical framework of fine-mapping. (3) Researchers have developed tools to joint analyze multiple traits to improve the power of gene discovery and to identify causal risk factors of diseases. Existing approaches, however, are often based on pair-wise analysis. We will develop a powerful statistical framework to better understand common biological processes driving genetic relationships among multiple traits. Additionally, we will develop more accurate Mendelian Randomization (MR) method to identify causal relationship among traits. (4) A key component of our effort is the development of user-friendly software that could benefit the broad psychiatric genetics community.

Public Health Relevance

Understanding genetics of neurodevelopmental and psychiatric diseases such as autism and schizophrenia will pave the way to better treatment of these complex and costly diseases. Sequencing or genotyping DNA of patients have discovered a number of genetic loci associated with the risk of these diseases, yet, to translate these data into knowledge of the underlying genes is challenging. The goal of this research is to develop a set of computational tools and software to interpret these large-scale genetic data from patients, facilitating the discovery of genes playing important roles in these diseases.

National Institute of Health (NIH)
National Institute of Mental Health (NIMH)
Research Project (R01)
Project #
Application #
Study Section
Genomics, Computational Biology and Technology Study Section (GCAT)
Program Officer
Arguello, Alexander
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Chicago
Schools of Medicine
United States
Zip Code
Liu, Yuwen; Liang, Yanyu; Cicek, A Ercument et al. (2018) A Statistical Framework for Mapping Risk Genes from De Novo Mutations in Whole-Genome-Sequencing Studies. Am J Hum Genet 102:1031-1047