Computational protocol: JAG: A Computational Tool to Evaluate the Role of Gene-Sets in Complex Traits

Similar protocols

Protocol publication

[…] JAG is written in the Python programming language and depends on PLINK [] and R []. JAG requires raw genotypic data as input files in PLINK binary format (.bim, .bed and .fam files) and an ASCII text file that contains the gene-set information. When summary statistics are used as input, JAG requires raw data from a reference population. JAG runs as a command line program on UNIX/Linux, Mac OS X and Windows operating systems. Typically, the use of JAG consists of three steps: SNP-to-gene annotation, self-contained testing, and competitive testing (see ). The first step may also be conducted outside of JAG. [...] We compared the performance of the self-contained gene and gene-set test of JAG with other tools that incorporate a self-contained test and use raw data as input.The performance of the self-contained gene-set test of JAG is compared with the performance of PLINK (v1.07), GATES (v2.5) and GRASS (v0.1). We used the default setting for each of these tools unless other settings were needed to accommodate comparison. For GRASS, we used the default settings except for the gene definition: using “abs” to map SNPs within the absolute genome location of the gene and a “dist” of “0” to indicate that SNPs are mapped to a gene only in case a SNP is located within the physical location of a gene (within TSS and TES). In addition, we conducted analyses in PLINK with a set of parameters that makes PLINK more comparable with JAG (--set-r2 1; --set-p 1; --set-max 99999).To compare the performance of the self-contained gene-based test in JAG we used VEGAS (v0.8.27), GATES (v2.5) and SETSCREEN (option in PLINK v1.07) with default settings. The performance of each of these tools is tested via a Type I error analysis and a power analysis. […]

Pipeline specifications

Software tools PLINK, VEGAS
Application GWAS
Diseases Crohn Disease