search

andreyshabalin / MatrixEQTL

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

Significant trans eQTLs are no longer significant when cis window increases #20

Closed katec04 closed 1 year ago

katec04 commented 1 year ago

Hi there,

I'm currently running MatrixEQTL to do methQTLs and I noticed something peculiar in my results.

Last week, I ran MatrixEQTL with separate cis and trans p-value thresholds. I used a cis p-value threshold of 1e-8 and a trans p-value threshold of 1e-14. I originally set a cis window of 10 kB. When I got my results back, there were 40,000 hits for trans, but zero hits for cis. I noticed that for a few trans hits, the associated SNP and methylation site were less than 50 kB apart. For one trans hit, they were 32 kB apart, for another, 22 kB, etc.

I reran MatrixEQTL with the same p-values, but I changed the cis window to 50 kB. I still have 0 cis hits. I am confused because my trans hits that were within 50 kB should now be cis hits, given I increased the cis window and the trans p-value is more stringent than the cis p-value.

I was wondering if you could help me understand why this might be occurring. Furthermore, if I ran MatrixEQTL without specifying a cis window, do you think this problem would be solved? Thank you in advance for your help!

andreyshabalin commented 1 year ago

Hi @katec04,

Would you check me$cis$ntests please. I wonder if your annotation files for genotypes and for methylation mismatch. For example, they could have different names for the same chromosomes (chr1 vs 1).

Andrey

katec04 commented 1 year ago

You're right - changing the chromosome naming solved the issue. Thanks!

On Thu, Oct 13, 2022 at 4:29 PM Andrey Shabalin @.***> wrote:

Hi @katec04 https://github.com/katec04,

Would you check me$cis$ntests please. I wonder if your annotation files for genotypes and for methylation mismatch. For example, they could have different names for the same chromosomes (chr1 vs 1).

Andrey

— Reply to this email directly, view it on GitHub https://github.com/andreyshabalin/MatrixEQTL/issues/20#issuecomment-1278198939, or unsubscribe https://github.com/notifications/unsubscribe-auth/A3TBAN7L6RIETMGQA4OB7CLWDB5L7ANCNFSM6AAAAAARETOQC4 . You are receiving this because you were mentioned.Message ID: @.***>