High % of reads lost to chimera filtering in DADA2

Hi all,

I was recently analyzing a data set of 16S V3V4 amplicons (from soil samples) and was running them through the DADA2 pipeline. I was losing 30-45% of my sequences in the chimera filtering step, and according to Dr. Callahan’s DADA2 tutorial, that’s way too high of a percentage.

One of my lab mates told me that she’s read that losing ~30% of your reads to the chimera filtering step in DADA2 and/or QIIME is normal when it comes to environmental microbiomes, but I am having a hard time finding that information in the literature.

My question is: if you’ve sequencing envrionmental microbiomes, have you lost a high % (>5% according to the DADA2 tutorial) of reads to the chimera filtering step? Is this common with environmental microbiomes as opposed to a gut microbiome (for example)?

Thank you for your input!