Data from: The genomic architecture of local adaptation in two connected populations of three-spined stickleback

Populations often harbor extensive genetic variation shaped by both
selection and connectivity, yet the genomic basis of this variation
remains incompletely understood. We generated two complementary datasets
in three-spined sticklebacks (Gasterosteus aculeatus) to characterize
single-nucleotide polymorphisms (SNPs) and structural variants (SVs)
across the genome. For SNP discovery, we used Illumina whole-genome
sequencing data aligned to the most recent reference genome, including the
Y chromosome for males, while excluding the pseudo-autosomal region to
minimize alignment errors. Reads were processed to remove duplicates,
clipped for overlaps, and locally realigned around indels, achieving an
average coverage of 11.2×. After standardizing coverage and removing
related individuals, we called SNPs chromosome by chromosome and filtered
for biallelic sites with a minor allele frequency > 0.05, coverage
between 4× and 35×, and a genotyping success rate exceeding 50%. To
complement this, we generated a structural variant dataset by combining
long-read (Nanopore) and short-read (Illumina) sequencing. Nanopore reads
>1 kb were mapped and filtered, and SVs were called independently
with multiple algorithms for each dataset to maximize confidence. Only SVs
detected by at least two callers per dataset were retained, and long- and
short-read datasets were then merged. Structural variants were genotyped
across all samples using a genome graph approach, with insertions
primarily resolved from long-read data and inversions largely from
short-read data. Together, these datasets provide a high-resolution view
of genomic variation, capturing both fine-scale SNPs and larger structural
variants. They enable the study of recombination landscapes, patterns of
differentiation, and the potential role of structural variation in local
adaptation across connected populations of three-spined sticklebacks.

种群往往蕴含着由选择作用与连通性共同塑造的丰富遗传变异，但这类变异的基因组基础仍未得到完全阐释。我们以三刺鱼（Gasterosteus aculeatus）为研究对象，构建了两套互补数据集，用于表征全基因组范围内的单核苷酸多态性（single-nucleotide polymorphisms, SNPs）与结构变异（structural variants, SVs）。针对单核苷酸多态性的发掘，我们采用了比对至最新参考基因组的Illumina全基因组测序数据，其中包含雄性的Y染色体，同时排除假常染色体区域（pseudo-autosomal region）以最大限度降低比对误差。对测序读段进行去重、重叠区修剪以及插入缺失（insertions-deletions, indels）区域的局部重比对处理，最终获得平均测序深度为11.2×的测序数据。在标准化测序深度并剔除近缘个体后，我们逐染色体进行单核苷酸多态性位点的分型，并筛选满足以下条件的双等位基因位点（biallelic sites）：次要等位基因频率（minor allele frequency）大于0.05、测序深度介于4×至35×之间、基因型分型成功率超过50%。为补充上述数据集，我们结合长读长（Nanopore）与短读长（Illumina）测序数据，构建了结构变异数据集。我们对长度大于1kb的Nanopore读段进行比对与过滤，并针对每个数据集采用多种算法独立识别结构变异，以最大限度提升结果可信度。仅保留每个数据集中至少被两个变异调用工具识别出的结构变异位点，随后将长读长与短读长数据集进行合并。我们采用基因组图谱（genome graph）方法对所有样本中的结构变异进行基因型分型，其中插入变异主要通过长读长数据解析，倒位变异则主要借助短读长数据完成。综上，这两套数据集可提供高分辨率的基因组变异图谱，同时涵盖精细尺度的单核苷酸多态性与大型结构变异。它们可用于研究三刺鱼连通种群的重组图谱、分化模式，以及结构变异在局部适应中可能发挥的作用。