RIVER is an
R package of a probabilistic modeling framework, RIVER (RNA-Informed Variant Effect on Regulation) that jointly analyzes personal genome (WGS) and transcriptome data (RNA-seq) to estimate the probability that a variant has regulatory impact in that individual. It is based on a generative model assuming that genomic annotations, such as the location of a variant with respect to regulatory elements, determine the prior probability that variant is a functional regulatory variant, which is an unobserved variable. The functional regulatory variant status then influences whether nearby genes are likely to display outlier levels of gene expression in that person. RIVER is trained in an unsupervised manner such that it does not require a labeled set of functional/non-functional variants; rather it derives its predictive power from identifying expression patterns that tend to coincide with particular rare variant annotations.
For more information about
RIVER, check the vignettes.
For further details of a list of genomic annotations used for constructing features and how to generate the features and outlier status, please refer to our submitted publication.
Get most recent version of R (>= 3.4) from CRAN.
## try http:// if https:// URLs are not supported source("http://bioconductor.org/biocLite.R") biocLite("RIVER")
Below is the citation output from using
citation('RIVER') in R. Please run this yourself to check for any updates on how to cite RIVER.
To cite the
RIVER package in publications use:
Li, X , Kim, Y , Tsang, EK , Davis, JR , Damani, FN, Chiang, C, Hess, GT, Zappala, Z, Strober, BJ, Scott, AJ, Li, A, Ganna, A, Bassik, MC, Merker, J, GTEx Consortium, Hall, IM, Battle, A , Montgomery, SB , "The impact of rare variation on gene expression across tissues", Nature 550, 239-243 (2017), doi:10.1038/nature24267, <URL: https://www.nature.com/articles/nature24267>.
Testing on Bioc-devel is feasible thanks to R Travis as well as Bioconductor's nightly build.