andrewGhazi/bayesianMPRA: A Bayesian framework for modelling MPRA data with an informative, annotation based empirical prior

This package provides a Bayesian modelling framework for MPRA data. Rather than using t-tests on derived quantities (i.e. log(mRNA/DNA) activities) or fitting linear models, this package employs a negative binomial model on the raw MPRA barcode counts. Implemented in Stan, it provides a posterior distribution on the transcriptional shift of variants. It bases the prior distribution on variant model parameters on empirical estimates based on genomic-annotation-based similarity.

install.packages("remotes")
remotes::install_github("andrewGhazi/bayesianMPRA")