Sequencing depth (coverage) can bias microbial intraspecies diversity estimates and how to account for it

Abstract

Unbiased estimation of intraspecific microdiversity is key for many microbiome studies. While it is well-appreciated that coverage (or sequencing depth) below 10X can provide biased estimates of microdiversity due to random sampling of alleles, the impact of variable coverage exceeding 10X on such estimates remains essentially unknown. Here we show that this impact is significant, even at coverage levels of > 100X, and provide approaches to effectively deal with this bias.