I want to use PICRUSt with some Illumina MiSeq 16S data that I’ve already processed on Mothur (using the MiSeq SOP), but I’m unclear about something. PICRUSt only works with Greengenes OTUs and I used the SILVA database when processing my data, so will I really have to redo the analysis (starting from the “pcr.seqs” step on SOP) in order to get this to work with PICRUSt? or is there a less time-consuming way?
No, you’d have to reclassify them (classify.seqs) usnig the greengenes reference. You might nag the picrust folks to see whether they can support the superior SILVA reference. FWIW, greengenes doesn’t seem to be getting updated any more.
Thanks for the response Dr. Schloss. That’s what I was thinking but I only have a taxonomy file for Greengenes, I’m not sure what I need to use for the reference parameter. When I ran this with SILVA initially I used the “.pcr.align” file I’d produced earlier as the reference. Basically my command line right now would look like this:
classify.seqs(fasta=BAC_R1.trim.contigs.trim.trim.good.unique.good.filter.unique.precluster.pick.fasta, count=BAC_R1.trim.contigs.trim.trim.good.unique.good.filter.unique.precluster.denovo.uchime.pick.count_table, reference=?, taxonomy=Gg_13_5_99.taxonomy.tax, cutoff=80)