martini is an R package to perform GWAS experiment that considers prior biological knowledge. This knowledge is modeled as a network of SNPs, were edges represent functional relationships between them (e.g. belonging to the same gene). Then, it looks for regions of the network associated with the phenotype using SConES.
martini like any Bioconductor package:
martini is a three step process:
library(martini) # 1. Read GWAS data with read.pedfile (or load the example :) ) data(minigwas) # 2. Create the SNP network: GS (structural information), GM (GS + gene # annotation information) or GI (GM + protein-protein interaction information) gs <- get_GS_network(minigwas) # 3. Find connected, explanatory SNPs (cones) cones <- search_cones(minigwas, gs) # cones$selected informs about whether the SNP is selected as cones or not head(cones) # snp chr cm pos allele.1 allele.2 c selected module # 3 1A1 1 0 10 A G 96.15385 TRUE 1 # 4 1A2 1 0 20 A G 96.15385 TRUE 1 # 5 1A3 1 0 30 A G 96.15385 TRUE 1 # 6 1A4 1 0 40 A G 96.15385 TRUE 1 # 7 1A5 1 0 50 A G 96.15385 TRUE 1 # 8 1A6 1 0 60 A G 96.15385 TRUE 1
Please, refer to the vignettes for more detailed usage examples.
martini results can be further examined using the blur package.