Fine-Mapping Tools : an interactive framework for dissecting disease-associated genetic loci with functional genomics data

Abstract

Fine mapping causal SNPs from GWAS summary statistics is hard. Although many frame- works exist to support fine mapping, some of which leverage epigenomic contexts to increase predictive power, they fail to provide interactivity. Here, we introduce Fine-Mapping Tools (fm-tools), a framework for doing interactive and iterative fine mapping. Fm-tools provides a harmonized data store and implements a number of algorithms for fine mapping -- one of which is the custom RiVIERA-mini, an efficient Bayesian inference framework -- and exposes them via a rich API that can be plugged into a variety of services (e.g., web applications for visualization). Most importantly, fm-tools allows scientists to interactively and iteratively explore dynamically generated hypotheses, as demonstrated by a case study for celiac disease. In summary, fm-tools standardizes the way fine mapping is done, reduces the overhead of fine mapping for scientists and of algorithm development for researchers, and paves the way towards achieving real-time personalized medicine.