order of commands in a pipeline: pre.cluster

Commands in mothur
1

Hi Pat + Forum,

I have a question regarding the order of commands for sequence processing, specifically the pre.cluster command.

In an older post (“chimera.slayer error”) you suggested the following option for a pipeline:
trim.seqs
unique.seqs
align.seqs
screen.seqs
filter.seqs
unique.seqs
pre.cluster
chimera.slayer

The order in the Costello pipeline is a bit different:
trim.seqs
unique.seqs
align.seqs
screen.seqs
filter.seqs
chimera.slayer
filter.seqs
unique.seqs
pre.cluster

In the first pipeline the pre.cluster commands is run before chimera.slayer where as in the second pipeline pre.cluster is run after checking for chimeras. I am wondering if it matters when pre.cluster is run (before or after chimera.slayer). And if so, why?

Any input or opinions would be great. Thanks!

Cheers,
V

2

Sorry for the confusion - the Costello example is correct. The main difference is really whether to do chimera.slayer before or after trimming so sequences overlap over the same alignment space. My logic for the new version is that it should be easier to detect chimeras from longer sequences, pre-trimmed than after.

3

Thanks!