Uncovering structural variation in conifer gigagenomes: evolutionary insights and technical challenges.

Abstract

Structural variants (SVs) are a major yet understudied source of genomic variation in conifers, whose large, repeat-rich genomes have hindered systematic SV discovery. Here, we combined whole-genome long-read and short-read sequencing to characterize the genomic landscape, functional impact, and evolutionary significance of SVs in a complex of three closely related pine species (Pinus densata, Pinus tabuliformis, and Pinus yunnanensis) with a hybridization history. From 21 long-read-sequenced individuals, we identified 5.7 million SVs, comprising 52% insertions, 43% deletions, and 5% inversions, duplications, and translocations. Approximately 97% of SVs were located in intergenic and intronic regions, and 60% overlapped transposable elements, whose activity shapes SV abundance and size variation. The proportion of loss-of-function (LoF) mutations was hundreds-fold higher among SVs than SNPs, with longer SVs more likely to cause LoF effects across all SV classes. Estimates of population diversity based on SVs and SNPs were largely concordant. In P. densata, the retention of parental SVs highlights the genomic signature of its admixed ancestry. We conducted graph pangenome-based SV genotyping in 29 short-read-sequenced individuals, yielding 44% recall and 70% precision, underlining the challenge of accurately recovering long-read-derived SVs in highly repetitive conifer genomes. Population-level selection scans on SNPs and genotyped SVs identified only 19% of candidate gene loci in common, indicating that the two marker types capture complementary components of environmental adaptation. Our findings demonstrate the importance of SVs as a dimension of genomic diversity and provide a foundation for integrating structural variation into evolutionary studies, conservation genomics, and tree breeding.