Core Hunter

Core Hunter generates diverse, representative core collections from large germplasm datasets to preserve genetic diversity and allele frequencies for plant breeding and genetic resource management.

Key Features:

• Multi-purpose subset selection: Employs local search algorithms to select core subsets based on distance metrics and allelic richness.
• Enhanced distance summarization: Introduces two novel methods for summarizing distances to improve diversity and representativeness relative to earlier versions.
• Algorithmic improvements: Implements a simple stochastic hill-climber and a parallel tempering algorithm to identify diverse cores and reduce variability across samples.
• Simultaneous optimization: Optimizes diversity together with representativeness or allelic richness and has been shown to outperform GDOpt and SimEli on these objectives.
• Versatility of input data: Supports genetic marker data, phenotypic trait data, or their combination for core construction.
• Performance and speed: Delivers improved effectiveness on diversity metrics and higher average and minimum distances between accessions with faster processing than prior versions.

Scientific Applications:

• Genebank curation: Enables reduction of genebank collection size while preserving genetic diversity and allele frequencies.
• Plant breeding: Produces representative and diverse subsets to support breeding program material selection and maintenance of genetic resources.
• Genetic and phenotypic studies: Facilitates selection of subsets for basic genetic analyses and applied agricultural research using marker and trait data.

Methodology:

Uses local search algorithms and distance metrics with two novel distance-summarization methods, optimizes allelic richness and representativeness via a stochastic hill-climber and parallel tempering, and performs simultaneous optimization of diversity, representativeness, and allelic richness using genetic marker and phenotypic data.