Shallow sequencing can mislead when evaluating hybridization capture methods

Abstract

Hybridization capture (also called enrichment) is a molecular technique used to focus high-throughput sequencing effort to a targeted region (or regions) of the genome, thereby reducing costs. This method has been used with great success in a wide range of research topics, including applied conservation genetics, and has been the subject of many studies aiming to optimize and improve the technique. In such technical optimization studies, it is common to minimally sequence (i.e., shallow sequencing) experimental capture libraries to save costs, and evaluate hybridization capture efficiency at lower sequencing effort than would be necessary for downstream analyses. Here I use simulations to show how this practice can produce misleading results and, ultimately, non-optimal experimental design choices. I argue that hybridization capture optimization studies must investigate their experiments at variable sequencing efforts and evaluate results in the context of the required target-space coverage for downstream analyses.