Abstract: Peanut (Arachis hypogaea) is an allotetraploid legume crop of agricultural and economic importance worldwide. Despite advances in peanut genomics, the integration of genetic diversity with multi-omics landscapes remains limited, and a comprehensive understanding of subgenome asymmetry is still lacking. Here, we present a high-quality genome assembly of an elite peanut cultivar Zhonghua 5 (ZH05; 2,622.85 Mb with 73,718 predicted genes) and construct a species-specific pan-genome based on 1,286 accessions, uncovering 849.82 Mb of non-reference sequences and 18,484 novel genes. Over 23 million genetic variants and extensive gene PAVs substantially expand the landscape of genetic diversity in peanut. Integrative multi-omics (transcriptome, DNA methylation, chromatin accessibility, and three-dimensional genome architecture), combined with population genomics analyses, reveal pronounced and multi-layered subgenome asymmetry. Specifically, SubA shows higher chromatin accessibility and overall gene expression, whereas SubB displays more transposable elements, elevated DNA methylation, more genes under selection during domestication, and stronger three-dimensional chromatin interactions with more stable topologically associating domains. Furthermore, genetic dissection using a recombinant inbred line population identified several major stable QTLs for seed size, whose pyramiding contributes to ZH05's superior yield performance. Collectively, this study establishes an integrated genomic framework for peanut, illuminates asymmetric subgenome evolution across sequence, epigenetic, and chromatin architecture, and provides valuable resources for accelerating molecular breeding and crop improvement.

Key words: Arachis hypogaea, genome assembly, multi-omics, pan-genome, QTL mapping, subgenome asymmetry