LDAK-PBAT is our tool for pathway-based association testing using summary statistics. In the preprint LDAK-PBAT: A Pathway-Based Analysis Tool for Decoding the Genetics of Complex Diseases (available on MedrXiv), we show that LDAK-PBAT is more powerful than the existing tools MAGMA and GSA-MiXeR.
Takiy, is there a better reference for magma (when used for pathway analysis)?
To run LDAK-PBAT requires GWAS summary statistics and a pathway tagging file. If you are analysing results from a human GWAS, it is easiest to use one of our pre-computed pathway tagging files (we provide versions suitable for analysing African, Asian and European GWAS results). Alternatively, we explain how to compute the tagging file yourself, for which you require a list of gene pathways and a suitable Reference Panel.
Always read the screen output, which suggests arguments and estimates memory usage.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Preparation:
First ensure that your summary statistics are in the format required by LDAK (see Summary Statistics for details). If you plan to use a pre-computed tagging file (which is our recommendation when analysing results from a human GWAS), we suggest you download the following file, and use it to calculate the overlap between your summary statistics and the HapMap3 SNPs.
Details of HapMap3 SNPs (All populations)
Be aware that the SNP names in your summary statistics should be rsIDs (e.g, names such as rs116720794 or rs3131972). If this is not the case, you may be able to use the above file (which contains genomic positions for both the hg19 and hg38 assemblies) to convert the current SNP names to rsIDs.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Download and extract the pre-computed tagging file:
Please note that if you are analysing non-human data, or would like to test your own gene pathways, you should instead Calculate Pathway Tagging.
Links here
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Test pathways for association:
The main argument is --sum-hers <outfile>.
This requires two options
--pathways <pathstem> - to specify the pathway tagging files.
--summary <sumsfile> - to specify the file containing the summary statistics.
Analyses can be sensitive to large-effect loci (particularly, if these are located in regions of extreme linkage disequilibrium, such as the MHC). These can be removed or moderated using the options --cutoff <float> and --truncate <float>; for example, --cutoff 0.01 will remove predictors that individually explain more than 1% of phenotypic variance, while --truncate 100 will reduce chi-squared test statistics above 100.
LDAK will report an error if the summary statistics file does not provide summary statistics for all predictors in the tagging file. If a relatively small proportion of predictors are affected (say less than 20%), it should be OK to override this error by adding --check-sums NO.
