Aromatic rice (Oryza sativa L. Indica) is popular among consumers. They provide significant economic benefit to producing countries as well as farmers. Aromatic germplasm is huge and diverse in terms of agro-morphological characters, grain types and aroma content. Development of a representative coreset of aromatic germplasm can be a good source for novel alleles identification and is required for its improvement. A collection of 552 aromatic accessions were screened for 40 DUS traits and 60 SSR markers. The data were processed in PowerCore software that implements the advanced M (or maximization) strategy using a modified heuristic algorithm and a coreset of 99 accessions was identified. The average Nei diversity and Shannon–Weaver diversity indices of the coreset were 0.635 and 0.823, representing the entire diversity of the collection. Statistical parameters like mean difference and variance difference were analysed to compare the mean and variance ratio between core and entire collections. Coreset has 3.45% mean difference and 14.4% variance difference. The coreset has 99.04% coincidence rate and 106.31% variable range which indicated the optimum properties of the coreset against the entire collection. This is the first attempt of making coreset of aromatic rice from larger germplasm.

香米（Oryza sativa L.Indica）在消费者中很受欢迎。它们为生产国和农民提供了巨大的经济利益。在农业形态特征，谷物类型和香气含量方面，芳香种质巨大而多样。芳香种质代表性核心组的开发可能是新等位基因鉴定的良好来源，并且是其改进所必需的。筛选了552个芳香种质，收集了40个DUS性状和60个SSR标记。数据在PowerCore软件中进行了处理，该软件使用改进的启发式算法实施了高级M（或最大化）策略，并鉴定了99个种质的核心集。核心集的平均Nei多样性和Shannon-Weaver多样性指数分别为0.635和0.823，代表了整个馆藏的多样性。分析了诸如均值差和方差之类的统计参数，以比较核心集合和整个集合之间的均值和方差比。核心集的平均差异为3.45％，方差为14.4％。核心集的重合率为99.04％，可变范围为106.31％，这表明核心集对整个集合的最佳性能。这是从较大的种质中制成香米核心的首次尝试。