Background and aims:
Vietnam harnesses a rich diversity of rice landraces adapted to a broad range of conditions, which constitute a largely untapped source of genetic diversity for the continuous improvement of rice cultivars. We previously identified a strong population structure in Vietnamese rice, which is captured in five Indica and four Japonica subpopulations, including an outlying Indica-5 group. Here, we leveraged on that strong differentiation, and the 672 rice genomes generated, to identify genes within genomic regions putatively selected during domestication and breeding of rice in Vietnam.
Methodology:
We identified significant distorted patterns in allele frequency (XP-CLR method) and population differentiation scores (FST), resulting from differential selective pressures between native subpopulations, and compared them with QTLs previously identified by GWAS in the same panel. We particularly focused on the outlying Indica-5 subpopulation because of its likely novelty and differential evolution.
Results:
We identified selection signatures in each of the Vietnamese subpopulations and carried out a comprehensive annotation of the 52 regions selected in Indica-5, which represented 8.1% of the rice genome. We annotated the 4,576 genes in these regions, verified the overlap with QTLs identified in the same diversity panel and the comparison with a FST analysis between subpopulations, to select sixty-five candidate genes as promising breeding targets, several of which harboured alleles with non-synonymous substitutions.
Conclusions:
Our results highlight genomic differences between traditional Vietnamese landraces, which are likely the product of adaption to multiple environmental conditions and regional culinary preferences in a very diverse country. We also verified the applicability of this genome scanning approach to identify potential regions harbouring novel loci and alleles to breed a new generation of sustainable and resilient rice.