haowulab / DSS

14 stars 13 forks source link

An error when running DMLtest #22

Open hengxing12 opened 2 years ago

hengxing12 commented 2 years ago

Hi

When I run my data, I meet an arror.

dmlTest = DMLtest(BSobj, group1=c("C1","C2"), group2=c("N2","N1"), smoothing=TRUE, BPPARAM=mParam)

Smoothing ... Estimating dispersion for each CpG site, this will take a while ... |======================================================================| 100%

|======================================================================| 100%

Error in split.default(1:n0, allchr2) : 组的长度为零但数据的长度大于零 Calls: DMLtest ... compute.mean.Smooth -> smooth.chr -> split -> split.default 停止执行

But if run: dmlTest = DMLtest(BSobj, group1=c("C1","C2"), group2=c("N1"), smoothing=TRUE, BPPARAM=mParam) It is ok.

dmlTest = DMLtest(BSobj, group1=c("C1","C2"), group2=c(""N2"), smoothing=TRUE, BPPARAM=mParam) it is ok

dmlTest = DMLtest(BSobj, group1=c("C1","C2"), group2=c("N2","N1"), BPPARAM=mParam) it is ok.

How can I do with this error?

Thanks,

hengxing

haowulab commented 2 years ago

This is strange. The Error in split.default seems to be caused by some chromosome has no CG site. Can you run the following and provide me the results:

seqnames(BSobj)

Better if you can share the BSobj so that I can try it. Maybe we can use dropbox for that if it's not overly big.

hengxing12 commented 2 years ago

Hi It is my fault. now is ok. thanks, hengxing

hongdoubao0315 commented 1 year ago

Hi It is my fault. now is ok. thanks, hengxing

Hi, I have a similar problem to you, if you remember, could you please provide your solution?