Error in quantile.default(sds, sd.threshold) : missing values and NaN's not allowed if 'na.rm' is FALSE

[vnapi...@gmail.com]

Hi

I am using Methylkit package for BSMAP methyratio output, for Rice Bisulfite data analysis, but when I am using the replicates for the two set of samples, error messages are coming like
Error in quantile.default(sds, sd.threshold) :
  missing values and NaN's not allowed if 'na.rm' is FALSE
please help me to find out the error in na.rm part, from where we can change this na.rm=True, is there any way to change this funciton from the code that we are using.
I am also getting error to analyse differential methylation of 4 sample set (2 control and 2 test). I am using this command

myDiff=calculateDiffMeth(meth)
save.image("/RELOAD2S/STAFF/SHARED/NGS_P2423/Methylkit_BSMAP/comparison_new/New_combination/Results/1vs13n2vs14/1vs13n2vs14.RData")
myDiff20p.hypo=get.methylDiff(myDiff,difference=20,qvalue=0.01,type="hypo")
myDiff20p.hyper=get.methylDiff(myDiff,difference=20,qvalue=0.01,type="hyper")
myDiff20p=get.methylDiff(myDiff,difference=20,qvalue=0.01)
write.table(getData(myDiff20p),file="/RELOAD2S/STAFF/SHARED/NGS_P2423/Methylkit_BSMAP/comparison_new/New_combination/Results/1vs13n2vs14/1vs13n2vs14_Diff20p.txt")
save.image("/RELOAD2S/STAFF/SHARED/NGS_P2423/Methylkit_BSMAP/comparison_new/Results/1vs13n2vs14/1vs13n2vs14.RData")


the Error it is showing is

Error in glm.fit(Tmod, X[!is.na(X)], weights = Y[!is.na(Y)], family = binomial(link = logit)) :
  object 'fit' not found
Calls: calculateDiffMeth ... glm.set.v1 -> t -> mapply -> <Anonymous> -> glm.fit
In addition: Warning messages:
1: In glm.fit(Tmod, X[!is.na(X)], weights = Y[!is.na(Y)], family = binomial(link = logit)) :
  no observations informative at iteration 1
2: glm.fit: algorithm did not converge
Execution halted


Please help me to solve the issue, its urgent for me.

Veena

[Altuna Akalin]

There are Probably NA values in your data. They could be coming from BSMAP, it is impossible to help without structured information. You should send a reproducible example with subset of your data reproducing the errors. 

--
You received this message because you are subscribed to the Google Groups "methylkit_discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to methylkit_discus...@googlegroups.com.
To post to this group, send email to methylkit_...@googlegroups.com.
Visit this group at https://groups.google.com/group/methylkit_discussion.
For more options, visit https://groups.google.com/d/optout.

--
Sent from mobile, excuse the brevity