Purchase this article with an account.
Elizabeth Rossin, Kasper Lage, Lucia Sobrin, Taibo Li, Mark Daly, Heiko Horn; Using protein-protein interaction analysis to understand genetics of heritable ocular disease. Invest. Ophthalmol. Vis. Sci. 2016;57(12):2637. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
We aim to build on methods we have previously published to analyze protein-protein interaction networks built from loci that emerge from genome-wide association studies (GWAS) of complex traits and disease. Interpreting GWAS is often challenging as it points to large loci that contain many genes, and the underlying culprit gene and the relationship among culprit genes are hard to discern. We have previously shown that protein-protein interaction networks can greatly aid in identifying underlying pathways implicated by genetic studies. Here we provide a significant upgrade to this method ("DAPPLE"). We apply it to age-related macular degeneration (AMD) to test the utility of the upgrade and to translate results of GWAS into an understanding of important genes and pathways.
Here we describe an approach to build and analyze networks of interacting proteins based on a previously published method called DAPPLE (disease associated protein-protein link evaluator). We have built a new underlying database of protein-protein interactions, expanding the available data 3-fold, and have applied new analysis methods to better understand complex network topography. We then build networks of interacting proteins that are encoded in loci associated to AMD and use a permutation approach to determine whether the network is significantly connected beyond chance expectation. By doing so, we are able to highlight potentially causal genes in large loci.
We apply the updated method to 19 published AMD loci. We find a strongly significant network underlying AMD that improves with the method upgrade (p<0.01) and not only involves the well established complement factor pathway but also pulls in additional loci. We therefore (1) provide a tool that is useful to anyone conducting large scale genotyping and sequencing studies and (2) show a highly significant network underlying the AMD loci that can be used to move from GWAS to a better understanding of biology.
Using information about protein-protein interactions has great power in translating loci that emerge from GWAS. We have previously developed early methods in the area of protein-protein interaction network analysis, and here we release a significant upgrade that is free for use by the community. Using this approach, we can point to specific genes and pathways that are implicated by large genetic studies.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.
This PDF is available to Subscribers Only