How can we sub.sample with higher iterations

Commands in mothur
1

Hi!
I was wondering for the sub.sample command how many iteration does it do to pick up randomly N number of sequence that we set for all samples to have the same number?
Is it possible to increase this iteration?

Thanks!

2

sub.sample only does one iteration. In the other commands like summary.single and dist.shared you can do it with the iter option.

Pat

3

Thanks Pat,

…But wouldn’t it be better to have more??? :roll:

4

For subsample, the goal is to generate a new shared file where every sample has been sub-sampled once for downstream use in things like lefse, metastats, etc. If we had more than one iteration, you’d need to run those downstream commands over and over as well. That’s probably a good practice, but you’d want to run all of those commands, not just sub.sample. If you ran sub.sample 1000 times and then got an average, that would effectively be the same as calculating the relative abundance on a shared file. The problem with this is that the data are no longer subsampled and samples with many more sequences would naturally have more OTUs, etc.

Make sense?
Pat

5

Thanks,
Trying to get my head around it! but yes I get what you mean!