Tian Ge, a PhD student in COMBI working with Prof Feng, will give an oral presentation (approx. 1% acceptance rate) at the Organization for Human Brain Mapping (OHBM) annual meeting to be held in Beijing in June, 2012.

Abstract of his talk titled 'Increasing Power for Voxel-wise Genome-wide Association Studies' is below:

The trend in imaging genetics is to embrace the brain-wide genome-wide association paradigm for nonrandom associations. To date, however, we are unaware of any method that has found associates that survive full-brain genome-wide correction for significance. This could be due to the mass-univariate nature of most methods, and how they do not have a biologically plausible way to model the interactions between SNPs. Here we present a suite of methods to improve speed and sensitivity of voxel-wise genome-wide association analyses while accounting for local genetic interactions. We find that the combination of our methods increases statistical power and produces significant associations using ADNI structural MRI data.