Search
Question: Use ChAMP to Find Differential Methylation Regions (DMR)
0
13 days ago by
yuabrahamliu0 wrote:

Hi there,

I always have a bug when use ChAMP to call differential methylated regions (DMR) and I have tried many times but I still cannot get over it, hope someone could give me some help.

I used ChAMP to analyze EPIC data and after Combat, I went to the step to find differential methylation probes (DMP) and regions (DMR). DMP could be called successfully, but DMR could not. My command is that,

myDMR <- champ.DMR(beta=myCombat2,pheno=myLoad_B$pd$Sample_Group,
arraytype = "EPIC", method="Bumphunter",cores=10)

I always met the same error, its information is,

Error in nulltabs[[i]] : subscript out of bounds
Calls: champ.DMR ... bumphunter -> bumphunter -> .local -> bumphunterEngine
Execution halted

Could anyone give me some help? Really appreciate you.

Best wishes,
Yu

written 13 days ago by yuabrahamliu0

Hello Yu:

champ.DMR() appeared few complain before, and I did not see this error much. I suspect that it could be a problem inside bumphunter, since ChAMP imported it's function, thus it would show here. See below link:

https://github.com/rafalab/bumphunter/issues/18

You may try adjust cores of parallel in your command. Seems doParallel package is a little bit outdated.

Best

Yuan Tian

Hi Yuan,

Your suggestion is very helpful! I reduced the number of cores required from 10 to 3, and then it worked smoothly! Although 3 cores will not be enough in the future because I will run on a large number of samples then, I am still very happy now.

Thank you so much Yuan!

Best wishes,

Yu