View Single Post
Old 12-17-2014, 10:51 AM   #4
n2hasan
Junior Member
 
Location: Denver, CO

Join Date: Sep 2009
Posts: 6
Default

Hello Jen,

I thought the Qiime workflow has an option to use datasets composed of reads from multiple regions of the 16s sequence, something analogous to the method implemented in Emirge and Emirge2, but I might be mistaken.

I've did a few initial analyses with the IonReporter. We're quantifying all the result numbers (Valid Reads, Ignored Reads, Reads Mapped, etc.) to better understand what the Ion 16s metagenomics workflow is actually doing with the data. So what I can say is that you have a few search options using the IonReporter: i) search against the LifeTech MicroSEQ curated database of ~15K full-length 16s rRNA sequences; ii) search against the Greengenes 16s rRNA database of ~400K full-length 16s rRNA sequence; iii) search using a two-step where it first searches reads against the MicroSEQ and searches reads that don't match MicroSEQ against the Greengenes database. It gives you OTUs based on a % identity threshold you specify when starting the workflow. Using default parameters for reads, as I understand, >97% identity results in genus level identifications; >99% identity results in species level identifications. It specifies slash calls as when reads have a tie in % identity range for two or more OTUs at a particular level (i.e. reads with 97% identity for both Mycobacterium and Nocardia). It compiles all the taxa represented across the different variable regions and estimates a consensus identification list of all the OTU identified with the different V region sequences detected in the sample. It doesn't appear that the IonReporter is trying to assemble the reads in any sort of way as is done with other workflows.

If people are looking for IonReporter to be a potential replacement for Qiime, SILVA, etc., then this is not the case. Those are capable of quite a bit more indepth analyses that are not part of IonReporter at this point.

My labmate and I are working this inbetween grant and paper edits, so progress is admittedly slow. I'll try to comeback as I find any new stuff.

Best Regards,

Nabeeh
n2hasan is offline   Reply With Quote