search

andreyshabalin / MatrixEQTL

Matrix eQTL: Ultra fast eQTL analysis via large matrix operations
53 stars 16 forks source link

Order of Matrix eQTL output #2

Closed QinqinHuang closed 6 years ago

QinqinHuang commented 6 years ago

Hi Andrey, Thanks for this amazing tool! In the output of Matrix eQTL, the most significant association ranks the first. When I was using it to map eQTLs in my simulated datasets, I found that some associations had different t-statistics but the same p-value (I think they may reach the minimum that can be calculated/displayed by R: 2.225074e-308). For these associations (or SNPs, they were for the same eGene), they are ordered by genomic locations. I was wondering whether it would make more sense if these SNPs could be sorted by t-statstics, which indicate significance levels. Best, Qinqin screen shot 2017-09-04 at 2 52 09 pm

andreyshabalin commented 6 years ago

Hi Qinqin, You are right, I hate zero p-values and I set them to .Machine$double.xmin in Matrix eQTL. I agree with you on the order of the output. It's an easy fix. I'll commit the update to Github tonight. Thank you. Andrey

andreyshabalin commented 6 years ago

I've committed the update. I'm closing this issue now.

QinqinHuang commented 6 years ago

Thanks Andrey!