Error in calculateDiffMeth with dispersion=MN

[alynn1993]

Hi, I am able to run calculateDiffMeth on my dataset when dispersion='none', but get the following error when dispersion='MN':
two groups detected:
will calculate methylation difference as the difference of
treatment (group: 1) - control (group: 0)
Error in if (is.na(p.value)) { : argument is of length zero
Calls: calculateDiffMeth ... .calculateDiffMeth -> simplify2array -> mclapply -> lapply -> FUN

Based on the calculateDiffMeth source code perhaps adding isTRUE to the if(is.na(p.value))) in the following could help:

do F-test when overdispersion >1 given
test=ifelse(test=="F" & phi>1,"F","Chisq")

p.value=switch(test,
F={
pf(deviance/phi, ddf, (length(w)-nprm), lower.tail = FALSE)
},
Chisq={
pchisq(deviance/phi, 1, lower.tail = FALSE)
})
if(is.na(p.value)) {
warning("Replacing NaN with 1.")
p.value <- 1
}

[alexg9010]

Hi @alynn1993,

Thanks for reporting this error.
Could you please send a minimal reproducible example to help us debug the problem?

Best,
Alex