Polygenic architecture of mental illness. Recent discoveries and future potential

The heritability for severe mental disorders is estimated to 60-80% and they are now regarded as complex genetic disorders, which are associated with the effects of multiple genes (i.e. ‘polygenic’) in combination with... [ view full abstract ]

The heritability for severe mental disorders is estimated to 60-80% and they are now regarded as complex genetic disorders, which are associated with the effects of multiple genes (i.e. ‘polygenic’) in combination with environmental factors. This makes it difficult to determine an individual’s risk, and to develop new treatments based on disease mechanisms. Genome-wide association studies (GWAS) have identified many trait-associated single nucleotide polymorphisms (SNPs). In psychiatric disorders only a small fraction of genetic variance has been identified despite recent large GWAS. This ‘missing heritability’ has been attributed to lack of proper statistical methods for analysis of the ‘polygenic architecture’.    

A large proportion of the ‘missing heritability’ is available within GWAS data when associations of SNPs are examined in aggregate. This implies the existence of numerous small genetic (‘polygenic’) effects that cannot be detected with traditional statistical methods. Thus, there is a need for innovative statistical approaches to enhance discovery of common variants. We have recently developed ‘enrichment tools’, a statistical framework for analyzing GWAS data building on Bayesian methods. Using summary statistics derived from SNP associations of large GWAS, functional genic elements showed differential contribution to phenotypic variance, with some categories (e. g. regulatory elements and exons) showing strong enrichment for phenotypic association.

Leveraging pleiotropy, i.e. genetic associations between two phenotypes from independent GWAS, the power for detecting small genetic effects is substantially increased, and novel susceptibility loci can be discovered. Applying these enrichment factors together in a new covariate modulated mixture model we identified several novel gene loci for schizophrenia. Of great interest is that many of the genes associated with the discovered loci are involved in neuronal excitability, including neurotransmitter systems (e.g., glutamatergic, dopaminergic, serotonergic, GABA-ergic, cholinergic) and ion channels (Na⁺, K⁺, Ca²⁺, and Cl^-). Thus novel Bayesian tools can increase power for uncovering more of the missing heritability (polygenic architecture) of psychiatric disorders, can increase discovery of gene loci that are more likely to replicate, and can improve risk prediction in new samples. Current work is focusing on improving these method and