Analyze NIH 16s data with DADA2

Question

[mb2subi]

I was wondering if is it possible to reanalize the 16s data from NIH (https://www.hmpdacc.org/HMR16S/) using DADA2 to infering ASV insted of OTUs. The main proiblem is that the files are fsa not fastq, so I don't know how to analyze using DADA2 or other ASV approach.

Any suggestion?

Answer 1

[M__]

The idea is here https://www.nature.com/articles/ismej2017119

In theory there is nothing wrong with your idea, its a good idea without question, however conversion between fsa and fastaq may not simplistic given the amount of data involved (it must be doable). I would assume that DADA2 rigidly sticks to fastaq?

I personally would strongly maintain adherance to OTUs.

[mb2subi]

I'm not very used to 16s data, but yes other ngs, is it possible to convert fsa to fastq using reformat.sh from BBMap?

Yes, at least I think, that DADA2 requires fastq input, perhaps someone more expert in this field can help us.

So, if I wasn't able to convert HMR16S to ASV is better to process my data as otus to compare with them, no?

Thanks a lot.

[M__]

This is just my opinion. Your idea involving ASV is good, despite people (like me) who are skeptical. That's what will sell in my opinion, its those ideas that win (where there is controversy and its new) and it doesn't have to be perfect. If there was extreme hostility publishing then I'd say OTUs.