High-quality genome assembly and pan-genome studies facilitate genetic discovery in mung bean and its improvement

Mung bean is an economically important legume crop species that is used as a food, consumed as a vegetable, and used as an ingredient and even as a medicine. To explore the genomic diversity of mung bean, we assembled a high-quality reference genome (Vrad_JL7) that was ∼479.35 Mb in size, with a contig N50 length of 10.34 Mb. A total of 40,125 protein-coding genes were annotated, representing ∼96.9% of the genetic region. We also sequenced 217 accessions, mainly landraces and cultivars from China, and identified 2,229,343 high-quality single-nucleotide polymorphisms (SNPs). Population structure revealed that the Chinese accessions diverged into two groups and were distinct from non-Chinese lines. Genetic diversity analysis based on genomic data from 750 accessions in 23 countries supported the hypothesis that mung bean was first domesticated in south Asia and introduced to east Asia probably through the Silk Road. We constructed the first pan-genome of mung bean germplasm and assembled 287.73 Mb of non-reference sequences. Among the genes, 83.1% were core genes and 16.9% were variable. Presence/absence variation (PAV) events of nine genes involved in the regulation of the photoperiodic flowering pathway were identified as being under selection during the adaptation process to promote early flowering in the spring. Genome-wide association studies (GWASs) revealed 2,912 SNPs and 259 gene PAV events associated with 33 agronomic traits, including a SNP in the coding region of the SWEET10 homolog (jg24043) involved in crude starch content and a PAV event in a large fragment containing 11 genes for color-related traits. This high-quality reference genome and pan-genome will provide insights into mung bean breeding.