Preprint on GWAS effects on biomedical research

By | February 7, 2017

Critics of the large investment made into genome wide association studies have noted that few GWAS results have had direct clinical impact. Supporters note that GWAS can have major indirect impact by revealing novel genes involved in a disease and motivating follow-up research. But this indirect impact has never been measured.

In our manuscript, we quantify the impact of GWAS on biomedical research into individual genes, through scientific publications. We first show that GWAS have not changed the historical bias of biomedical research toward genes involved in Mendelian disease. We then show that GWAS hits do receive additional publications compared to control genes and model expectations, but this impact is diminishing. Our results thus demonstrate the need for reforms to ensure that promising GWAS hits are pursued.

Paper on network dynamics and protein evolution published

By | July 5, 2016

Congratulations to Brian Mannakee on his paper that just came out in PLoS Genetics. In it, we use dynamical systems biology models to predict rates of protein domain evolution. The success of these predictions suggests that natural selection is acting to preserve network dynamics in the face of perturbations from mutations. Although the models are imperfect, we think this work shows the power of systems biology models for quantitatively predicting evolutionary processes.

journal.pgen.1006132.g001

Paper on triallelic population genomics published

By | March 30, 2016

Triallelic figureCongratulations to Aaron Ragsdale on his paper that was just accepted for publication in Genetics. In it, we develop a novel diffusion model for trialleic sites, which required some non-trivial applied mathematics. We then apply the model to mutually nonsynonymous codons in Drosophila, to infer the correlation between selection coefficients for mutations at the same protein site. Remarkably, our inferred correlation agrees quantitatively with direct biochemical experiments in bacteria and yeast, suggesting that the correlation we measure is a fundamental property of protein evolution.