Zhiyao Ma, Xiaodong Xu, Wenjing Peng, Tianhao Zhang, Zhuyifu Chen, Shuo Cao, Fan Zhang, Yiwen Wang, Hua Xiao, Yingchun Zhang, Zhenya Liu, Zhongjie Liu, Hui Xue, Qiming Long, Ting Hou, Wenrui Wang, Yuting Liu, Zhongxin Jin, Mengya Zhang, Yanmei Peng, Jun Wen, Brandon Gaut, Yongfeng Zhou
Amur grape (Vitis amurensis Rupr.) is widely recognized for its cold tolerance traits and serves as a valuable genetic resource for breeding climate-resilient grape cultivars. Here, we construct a graph pangenome reference (Vampan_V1.0) and generate a variant map comprising 48,308,434 short variants and 127,094 TE-associated structural variants (TEVs) using deep resequencing data from 330 samples across 31 natural populations covering the species’ distribution range. We discover a biased accumulation of SNPs around TEVs and identify 823 candidate adaptive genes associated with environmental variables. Using machine learning-based genetic offset models, we further show that putative adaptive TEVs significantly reduce genetic offsets by 7.3% to 8.2% under future climate scenarios. Our study shows the power of a graph-based pangenome to resolve complex variation and highlights the impact of TEVs on genetic diversity, local adaptation, and resilience to future climate change, providing insights into utilizing crop wild relatives in climate-resilient crop breeding. Amur grape (Vitis amurensis) is a hardy vine known for its resilience in cold climates. Here, the authors construct a phased graph-based pangenome based on the eight haplotype-resolved genomes and reveal the association of transposable elements-linked variants with climate adaptation.