hruffieux/atlasqtl: Flexible sparse regression for hierarchically-related responses using annealed variational inference

Flexible sparse regression for variable selection in large predictor and response settings, based on a series of hierarchically-related spike-and-slab submodels. The model is also tailored to the detection of hotspots, namely, predictors associated with multiple responses, which it represents using a global-local horseshoe specification. Inference uses closed-form variational updates coupled with a simulated annealing algorithm to enhance exploration of highly multimodal spaces. This software allows joint inference at large scale, e.g., for molecular quantitative trait locus (QTL) studies, where the hotspots are genetic variants regulating several molecular traits simultaneously. See H. Ruffieux, A. C. Davison, J. Hager, J. Inshaw, B. Fairfax, S. Richardson, and L. Bottolo. A global-local approach for detecting hotspots in multiple response regression. The Annals of Applied Statistics, 14:905-928, 2020.

install.packages("remotes")
remotes::install_github("hruffieux/atlasqtl")