mani2012/BatchQC

Error in if (var(data.matrix[i, cond2[[j]]]) == 0) { : missing value where TRUE/FALSE needed

widsquid opened this issue · 3 comments

Hello, thanks for the great package. I successfully went through the demo but am having some trouble with my data.

batchQC(counts, batch = batch, condition = cond, interactive = TRUE)
Error in if (var(data.matrix[i, cond2[[j]]]) == 0) { :
missing value where TRUE/FALSE needed

One clue is that my many of my conditions have only one replicate. Does this only work if there are multiple reps per condition (for all conditions)?
Thanks so much.

Ahh yes of course, thanks so much for the reply. Is there any way around this? For example, if I am trying to normalize for batch effect across ~600 RNA-seq samples that were collected from a series of 8 experiments (batches) where each condition is present only once with the exception of 8 (pos and neg ctrl) conditions which were replicated in each batch. Is there any way to use the information from the 8 samples that appear across all batches to extrapolate a batch correction to the entire set of conditions? Thanks again.
W