Description of the package
The PINTS (Protein Interaction Network Tissue Search) package provides a framework to identify groups of interacting genes with high user--specified scores and perform tissue--specific expression enrichment for any significant groups identified. We have used it to demonstrate that genes under purifying selection in the human population are clustered, suggesting selection acts on entire biological mechanisms.
You can install the PINTS package from github.
install.packages('devtools') library(devtools) install_github("CotsapasLab/PINTSv1") library(PINTS)
In addition to this, you would need to install all package dependencies. They are downloadble frm CRAN (http://cran.r-project.org) and Bioconductor (http://www.bioconductor.org) project website. The following are the necessary packages.
#Installing from CRAN install.packages("igraph") #installing from Bioconductor.org source("http://bioconductor.org/biocLite.R") biocLite("BioNet") biocLite("RBGL") biocLite("graph") biocLite("Matching")
For the tissue-level context search in the step 3 of the PINTS workflow, you also need to have a UGM software running under MATLAB. The UGM software is available from http://www.di.ens.fr/~mschmidt/Software/UGM.html. We provide a set of Matlab scripts used for PINTS workflow separately (Download inst/extdata/UGM_2011_CotsapasAdd.zip and upzip the file under a working direcotry).
A series of tutorials is intended to demonstrate the PINTS workflow to identify the mutational constraint gene subnetwork and to analyize the biological context (tissue specificity) of the identified disease-associated subnetwork.
You can check out the tutorials in a R session after the installation of the package.
Please cite JinMyung Choi et al, Network Analysis of Genome-Wide Selective Constraint Reveals a Gene Network Active in Early Fetal Brain Intolerant of Mutation, PLoS Genetics 12(6):e1006121 2016.