Relative performance of two DNA extraction and library preparation methods on archaeological human teeth samples

Abstract

DNA extraction and library preparation are crucial steps in any ancient DNA study. Although palaeogenomic researchers are facing a growing choice of DNA extraction and sequencing library preparation methods, how their performance varies with DNA preservation remains unclear. To help elucidate this question, we compared the performance of two common DNA extraction and Illumina library preparation methods on a set of archaeological human samples, considered to contain ancient DNA of intermediate to good preservation (5–50% endogenous DNA). Results indicate that while the levels of contamination and endogenous DNA recovered are comparable for both silica-in-solution and silica-column based extractions, the ability of the former to accommodate larger starting quantities of sample material confers notable benefits with regards to library complexity, and furthermore seems to aid with the recovery of shorter endogenous DNA molecules. While our observations gained from comparing the single-stranded with double-stranded DNA library construction methods largely replicate earlier observations, the combination of our data with previously published datasets demonstrate that the benefits gained using single-stranded methods are inversely proportional to the endogenous DNA content in the ancient sample.