Hello, in Camarinha-Silva et al. 2014, the mean number of sequences per sample was 7682 ± 484, totalling 1 459 615 usable paired-end sequence reads. In your study, after quality filtering, a total of 782,683 paired-end reads, with an average of 41,194 reads per sample (a minimum of 30,101 reads and a maximum of 62,232 reads) were obtained. So your average reads is >5 folds more than previous study. So my question is do you think that it will affect your results if your average reads is about 7600? That will decrease the experiment cost, won't it?

Sincerely,

Jack Yang