CRISPR host signal reduction for metabarcoding

Metabarcoding-based methods for identification of host-associated eukaryotes have the potential to revolutionize parasitology and microbial ecology, yet significant technical challenges remain. In particular, highly abundant host reads can mask the presence of less-abundant target organisms, especially for sample types rich in host DNA (e.g. blood, tissues). Here we present a new CRISPR-Cas9 mediated approach designed to reduce host signal by selective amplicon digestion, thus enriching clinical samples for eukaryotic endosymbiont sequences during metabarcoding. Our method achieves a nearly 76 % increased efficiency in host signal reduction compared to no treatment and a nearly 60 % increased efficiency in host signal reduction compared to the most commonly used published method. Furthermore, application of our method to clinical samples allows for detection of parasite infections that would otherwise have been missed.