Data from: High-throughput SNP genotyping of historical and modern samples of five bird species via sequence capture of ultraconserved elements

Sample availability limits population genetics research on many species,
especially taxa from regions with high diversity. However, many such
species are well represented in museum collections assembled before the
molecular era. Development of techniques to recover genetic data from
these invaluable specimens will benefit biodiversity science. Using a
mixture of freshly preserved and historical tissue samples, and a sequence
capture probe set targeting >5000 loci, we produced high-confidence
genotype calls on thousands of single nucleotide polymorphisms (SNPs) in
each of five South-East Asian bird species and their close relatives (N =
27–43). On average, 66.2% of the reads mapped to the pseudo-reference
genome of each species. Of these mapped reads, an average of 52.7% was
identified as PCR or optical duplicates. We achieved deeper effective
sequencing for historical samples (122.7×) compared to modern samples
(23.5×). The number of nucleotide sites with at least 8× sequencing depth
was high, with averages ranging from 0.89 × 106 bp (Arachnothera, modern
samples) to 1.98 × 106 bp (Stachyris, modern samples). Linear regression
revealed that the amount of sequence data obtained from each historical
sample (represented by per cent of the pseudo-reference genome recovered
with ≥8× sequencing depth) was positively and significantly (P ≤ 0.013)
related to how recently the sample was collected. We observed
characteristic post-mortem damage in the DNA of historical samples.
However, we were able to reduce the error rate significantly by truncating
ends of reads during read mapping (local alignment) and conducting
stringent SNP and genotype filtering.