Scalable Methods for Understanding the Genetic Architecture for Complex Traits
Thanks to the decreasing cost of sequencing and improved statistical methods for genotype imputation, it is now possible to aggregate large datasets with millions of individuals to study the genetics of complex traits. Previous studies on the smoking and drinking addictions were hampered by the lack of power and very few genes were identified in genome-wide association studies. Recently breakthrough in human addiction genetics were made possible by large datasets, where 406 new loci for smoking/drinking addiction were identified. In this talk, I will discuss scalable meta-analysis methods that enable the association analysis of this large datasets, methods for fine mapping causal genetic variants and methods for assessing the replicability of the identified association signals. The biological and clinical implications of the identified genetic associations will also be discussed.