SimEli: Similarity Elimination Method for Sampling Distant Entries in Development of Core Collections

ABSTRACT Sampling core collections containing a diverse set of entries has been practiced over the last two decades for a number of crops and has become a vital component of modern day crop improvement programs. A diverse, multipurpose core collection should represent the maximum genetic diversity available in an entire germplasm collection with a small number of entries. Selection of genetically distant entries that represent the maximum diversity of the entire germplasm collection is a challenging task that has been improved over the years. In this study, we introduce the similarity elimination (SimEli) method to sample genetically distant entries for the development of core collections, which was used to sample a diverse core collection of mulberry accessions using phenotypic markers. The performance of the SimEli method was compared with that of the PowerCore algorithm for phenotypic markers and with that of the Core Hunter and genetic distance optimization (GDOpt) algorithms for simple sequence repeat (SSR) markers. The SimEli method effectively selected genetically distant entries, whereas PowerCore proved efficient for selecting outliers among a small number of entries. However, the SimEli method outperformed the Core Hunter algorithm in selecting distant entries with high mean and minimum entry to nearest entry distance values. The Core Hunter collections retained a greater number of alleles than did collections developed using the SimEli method only when increased weight was given to Shannon's diversity index when using Core Hunter. The SimEli method is more user‐friendly, involves simple steps, and requires less computational time than other leading programs for the development of core collections.