jennasimit/MTFM: Multinomial fine-mapping

MFM is a computationally efficient approach to simultaneously fine-map several related diseases and borrows information between diseases in a Bayesian framework. Joint models that have at least one shared causal variant between a pair of diseases are given higher weight than those without shared variants. MFM has greater accuracy than single disease analysis when shared causal variants exist, and negligible loss of precision otherwise. The best SNP models are summarised in terms of sets of correlated SNPs that have a similar impact on the trait. This approach is an extension of the GUESSFM software for fine-mapping of single traits and relies on several of the functions from that software.

install.packages("remotes")
remotes::install_github("jennasimit/MTFM")