Abstract
BACKGROUND: Generating parasite genomes is challenging when little of the DNA in infected host tissue is from the parasite. We used selective whole genome amplification (SWGA) to generate genomic data from wildlife samples of the avian haemosporidian Haemoproteus majoris (lineage PARUS1) and its host, the blue tit (Cyanistes caeruleus). METHODS: We used SWGA to amplify the parasite DNA in nine avian blood samples collected between 1996 and 2021, and subsequently performed short-read sequencing and bioinformatically separated the host and parasite reads in each sample. RESULTS: SWGA increased the percentage of parasite reads significantly. Sequencing to a depth of about 56 million reads (forward and reverse) per sample resulted on average (± standard error [SE]) in 11.3X ± 1.85 for the host genome and 1.17X ± 0.446 mean depth of coverage for the host and parasite, respectively, after SWGA. Furthermore, about 74% of the host genome (genome size approx. 1.2 Gb) and 33% of the parasite genome (approx. 23.9 Mb) had at least 1X coverage on average; two samples had 1X coverage of approximately 60% of the parasite genome. Parasite sequencing success was positively correlated with parasitemia. When comparing the parasite sequences in the four best samples, we identified 9895 sites (minimum 5X coverage) that varied among the infections. When filtering the full dataset to at least six samples per variant, we identified 14,512,339 and 7068 sites that varied among samples in the host and parasite populations, respectively, revealing variation among samples and years. CONCLUSIONS: SWGA facilitates dual host-parasite population genomics in this system and will greatly expand our understanding of host-parasite interactions over space and time.