count.groups

Hi,
I am using the MiSeq SOP on a set of 18S. I have been playing with some of the parameters, and at some point on one of my trials I lost 99% of my reads, for a fifth of my samples. In order to check where this happen I thought about running count.groups() but the command does not always read my .group files.
Anything I am not getting here?
Any other idea to keep checking the reads per sample (group) I am creating?

THANKS!

Carla

You should probably use summary.seqs to count the total number of sequences at each step.

Pat