Could "chloroplast" sequences actually be mitochondrial?

[xchromosome]

Hello all!

I'm studying the gut microbiota of chickens. I have jejunal contents/digesta and also mucosal scrapings. See attached/below the taxonomic assignments (using SILVA) for the taxa in each of my treatment groups! (AGP is antibiotic-supplemented).

My question is... there's a lot of "Cyanobacteria" here, which, if you look at the genus level assignments, turn out to be mostly (99.1%) chloroplasts. BUT... I'm suspicious of this because of the high relative abundance of them in the mucosal scrapings in particular. Surely if you were expecting chloroplasts (presumably from the chicken feed), you'd get them in the gut digesta. And you do get some, but the mucosa have a much high level, which you wouldn't really expect. Those scrapings DO, however, inevitably contain a lot of chicken DNA from scaping the gut. Those samples contained a very high concentration of DNA but were difficult to amplify with my 16S primers and I suspected that was because a lot (most?) of the DNA was chicken rather than bacterial.

I know the mitochondria and chloroplasts are both closely related to bacteria and I've already eliminated them from my table before diversity analysis etc. I'll be manually taking them out of this graph too, but I was just wondering if it seems likely that the "chloroplasts" in this case are actually mitochondrial sequences and how I might go about confirming that?

Thanks so much!

[SoilRotifer (Mike Robeson)]

Hi @xchromosome,

These really might be chloroplast sequences, as they are quite different from mitochondrial sequences. I agree, depending on chickens' diet / feed, you might be amplifying food items from their gut. Assuming their food has a plant component. You can confirm this in a few ways:

• Take your suspect features (sequences) and submit them to the SILVA-ACT (Alignment, Classification and Tree Service). If you select all the appropriate options, this tool will return an alignment and tree visualization of your sequences against their nearest neighbors. You need at least 4 sequences for this to work. If these are indeed chloroplasts then they should group with other chloroplast (e.g. plant species) sequences to the exclusion of bacteria / cyanobacteria .
• Downloading several well-annotated and trusted chloroplast 16S rRNA sequences, and cyanobacterial sequences, either from GenBank or SILVA, and place them in a file with your ASVs. Then construct an alignment and phylogeny. Then observe where your sequences fall, I'd suspect they'd fall in with the true chloroplast sequences to the exclusion of the cyanobacterial sequences.
• BLAST the suspect sequences via the QIIME visualization or manually through GenBank.

-Cheers!
-Mike