Comparing The Impact of Reference Genome Divergence and Assembly Level on Population Genomic Analyses of Exome Capture Data

The reference genome is a comprehensive representation of an organism's DNA sequence, used in population genomics for aligning sequencing reads and detecting genetic patterns across samples. For species with large and complex genomes, such as those in the genus Pinus, generating high-quality references has proven challenging. Until 2022, the only available reference was from the North American Pinus taeda. However, phylogenetic distance between the reference and target species can introduce biases in polymorphism detection. With the recent release of a reference genome for the Chinese pine (Pinus tabuliformis), European Pinus species now benefit from a closer and higher-quality genomic framework. In this study, I assess how reference genome divergence affects population genomic inference in exome-capture data from two species: the Maritime pine (Pinus pinaster) and the Scots pine (Pinus sylvestris). To investigate this, I compare the effects of each reference genome to each of the two species across SNP discovery, filtering, population structure, summary statistics, demographic inference, and genome scans. The new reference genome significantly influences all analyses, with more pronounced differences in P. sylvestris, the species most closely related to the new reference genome. Specifically, it reveals clearer population structure and novel insights into genetic diversity for P. sylvestris, and more accurate results on demography, summary statistics estimates and selection signals in both species. Our main finding is that the choice of reference genome alters SNP frequency distributions, which in turn affects most downstream analyses. This study highlights the extent and direction of biases introduced when using suboptimal reference genomes for mapping sequencing reads and SNP calling in population genomic analyses.