Effects of sample age on data quality from targeted sequencing of museum specimens: what are we capturing in time?

Abstract

Background: Next generation sequencing (NGS) can recover DNA data from valuable extant and extinct museum specimens. However, archived or preserved DNA is difficult to sequence because of its fragmented, damaged nature, such that the most successful NGS methods for preserved specimens remain sub-optimal. Improving wetlab protocols and comprehensively determining the effects of sample age on NGS library quality are therefore of vital importance. Here, I examine the relationship between sample age and several indicators of library quality following targeted NGS sequencing of~1300 loci using 271 samples of pinned moth specimens (Helicoverpa armigera) ranging in age from 5 to 117 years.Results: I find that older samples have lower DNA concentrations following extraction and thus require a higher number of indexing PCR cycles during library preparation. When sequenced reads are aligned to a reference genome or to only the targeted region, older samples have a lower number of sequenced and mapped reads, lower mean coverage, and lower estimated library sizes, while the percentage of adapters in sequenced reads increases significantly as samples become older. Older samples also show the poorest capture success, with lower enrichment and a higher improved coverage anticipated from further sequencing. Conclusions: Sample age has significant, measurable impacts on the quality of NGS data following targeted enrichment. However, incorporating a uracil-removing enzyme into the blunt end-repair step during library preparation could help to repair DNA damage, and using a method that prevents adapter-dimer formation may result in improved data yields.

Predictors of sequence capture in a large-scale anchored phylogenomics project

“…Short-read next-generation sequencing methods require short fragments of DNA and can sequence DNA from old specimens. Thus, NGS has become a common alternative to PCR and Sanger sequencing, enabling incorporation of museum and herbarium samples in projects that require DNA sequencing (McGaughran, 2020;Mayer et al, 2021;Raxworthy and Smith, 2021). However, severe degradation that produces fragment lengths below the target length of the sequencing method will likely prevent a sample from being captured.…”

Unlocking inaccessible historical genomes preserved in formalin

“…However, without a framework to guide specimen selection, genomic work on formalin-preserved museum tissues will remain infeasible. It is probably impossible to fully know the numerous and interdependent factors driving sequencing success, for example, age of the specimen (McGaughran, 2020;Watanabe et al, 2017), method of preservation (Zimmermann et al, 2008), postmortem interval (Bär et al, 1988) and heat and light exposure during storage. However, identification of metrics with which to prescreen specimens for sequencing suitability will improve yield of genomic data while reducing unnecessary destruction of specimens.…”

Retrospective genomics highlights changes in genetic composition of tiger sharks (Galeocerdo cuvier) and potential loss of a south-eastern Australia population

“…Our findings are unlikely to be the result of technical issues associated with the use of historical DNA (hDNA). In general, capture sequencing of hDNA has a lower sequencing depth and coverage leading to fewer mapped reads and thus more missing data with increasing sample age 47 , both of which could affect downstream population genomic inferences 48 – 50 . Although our historical samples presented fewer reads than the contemporary samples, the level of missing data was generally low, even for the oldest samples (maximum level of missing data allowed was 5% per SNP).…”