Hello, thank you for developing such a user-friendly eqtl analysis software.
My data contains 7,788,068 SNPs, which is quite a lot, and there are around 180 samples.
I have two questions.
First, I noticed that when running --mode cis, you can use the distance parameter --cis-window. When running --mode trans, does the software correlate all SNPs with all gene expression levels? I previously ran trans, and it defaulted to outputting p-values less than 1e-5. If I run cis and trans separately, how can I ensure that the p-values of cis and trans are on the same scale? This will affect subsequent eqTL clustering and selection.
Second, to obtain association results on a uniform scale, I'm currently using --mode gwas to correlate all SNPs with genes. I've checked, and the current results are more in line with expectations, but on a single server, the computational load is too high and it's slow. Do you have any better suggestions?