README.md

In sahildshah1/gene-surrounder: GeneSurrounder

Network-based identification of disease genes in expression data: the GeneSurrounder method <!-- Sahil Shah sahil.shah AT u.northwestern DOT edu

--- -->

The GeneSurrounder package implements the method we previously developed to identify disease- associated genes from expression data and an independent network model of cellular interactions. We developed GeneSurrounder to find the genes with neighbors on the network that are differentially expressed (with the magnitude of the differential expression decreasing with distance from the putative disease gene) and have correlated expression with the putative disease gene. Since the differential expression of the neighbors of a putative disease gene does not depend on their association with that gene, our algorithm consists of two tests that are run independently of each other. Their results are then combined to determine if the putative disease gene is a central candidate disease gene.

In order to illustrate our method, we apply our algorithm to one study of high- vs-low grade ovarian cancer from the publicly available and curated collection curatedOvarianData (GEO accession GSE14764). We have constructed the global network model from KEGG pathways. To follow the example:

1. Clone this repo, change directories to gene-surrounder/vignettes/figs, and open vignette-main.pdf
2. Open a R interpreter, change directories to gene-surrounder/vignettes/figs, and follow the example in vignette-main.pdf

sahildshah1/gene-surrounder documentation built on May 27, 2019, 7:27 a.m.