This page provides detail on installing and running a complete SNP effect concordance analysis (SECA) using genome-wide association study (GWAS) summary results (Nyholt 2014).
The ‘SECA_local_version.zip’ download contains the scripts necessary to run a SECA analysis on your local linux computer. As detailed in the ‘local_SECA.README’ file, the scripts assume (require/expect) the input files to contain independent SNPs (i.e., LD clumped data as required by the iSECA webtool).
I have also placed the original SECA ‘SECA_ SNP effect concordance analysis.html’ and iSECA ‘iSECA_ SNP effect concordance analysis.html’ webpages in the ‘SECA_local_version.zip’ download for your reference.
The SECA local version requires input from independent SNPs, therefore users must perform their own linkage disequilibrium (LD) clumping. Users should refer to my LD clumping tutorial (archived version).
On the LD clumping tutorial webpage I’ve included numerous PLINK (binary) format genotype data file downloads for use in the LD clumping. For most cases, you will want to use the ‘1000G_20101123_v3_GIANT_chr1_23_minimacnamesifnotRS_CEU_MAF0.01.zip’ dataset to LD clump SNPs using LD estimates in a sample of European Ancestry with rsIDs as the SNP names. Note, multiple other genotype data files (including for different ancestries) are also available on the tutorial webpage.
The instructions are quite straight forward and hopefully the perl script and compiled C++ program (‘gwama2.1_SECAalign’) will run on your linux system.
Briefly, the ‘local_SECA.README’ file details how to check the file formats (perl script), align the SNP effects (using the ‘gwama2.1_SECAalign’ program) and perform the various statistical analyses using the R language and environment for statistical computing and graphics (R scripts). If the compiled ‘gwama2.1_SECAalign’ program will not run on you system, you can try compiling your own version using the source files in gwama2.1_SECAalign_source.zip.
Remember, you need to first perform LD clumping prior to running the R scripts.
Follow these links to view archived versions of the original SECA and iSECA homepages.