Research Highlights

Computer model turns gene loci into disease mechanisms

Published online 5 February 2014

Biplab Das

While genome-wide association studies (GWAS) have proved invaluable in uncovering genetic loci associated with diseases, pinning down the exact disease causing variants within these loci and how a variant can cause disease remains a challenge. Now, an international research team from Germany, Qatar, the United States and Norway has developed a computer-based model called phylogenetic module complexity analysis (PMCA) to track disease-causing genetic variations in the human genome.

Using PMCA, the researchers identified genetic variations in non-coding genome regions that bind transcription factors to regulate gene expressions, showing how they can be linked to disease mechanisms, and published their findings in Cell.

To test PMCA, the researchers applied it to a set of 47 risk loci for type 2 diabetes (T2D), causing complications such as insulin resistance and impaired insulin secretion in computer simulation studies. They found gene-regulating variations in genomes resulted in 3.1- to 101-fold increase in DNA-protein binding.

To test their findings experimentally, the researchers zeroed in on a single-base variation in the non-coding region of PPARG gene in human adipose stromal cells. This variation blocked the activity of PPARG2 protein, which plays a role in fatty acid storage, disrupting glucose metabolism and initiating T2D.

They suggest that PMCA can be used in various diseases such as cancer, myocardial infarction and thyroid hormone resistance. "The results offer great promises for developing better diagnostic means and tailored therapies for various genetic disorders," says Melina Claussnitzer, a co-author of the study.


  1. Claussnitzer, M. et al. Leveraging cross-species transcription factor binding site patterns: from diabetes risk loci to disease mechanisms. Cell. 156, 343-358 (2014) doi: 10.1016/j.cell.2013.10.058