Structure Harvester & STRUCTURESelector

Introduction

Understanding population structure is a cornerstone of genetics, plant breeding, and evolutionary biology. The STRUCTURE software provides a Bayesian clustering method to assign individuals to populations based on multilocus genotype data. However, interpreting results and identifying the optimal number of clusters (K) can be challenging. Tools like Structure Harvester and STRUCTURESelector have become essential for analyzing and visualizing STRUCTURE output.

What is Structure Harvester?

Structure Harvester is a web-based and standalone program designed to visualize STRUCTURE results. It implements the popular Evanno method for determining the most likely number of genetic clusters (K). The tool generates clear plots of likelihood values and ΔK, helping researchers identify meaningful population partitions.

How It Works

Structure Harvester processes the results_summary.txt files produced by STRUCTURE. It extracts log-likelihood scores (LnP(D)) and applies the Evanno ΔK method (Evanno et al., 2005). This allows researchers to pinpoint the value of K that best explains population structure.

Step	Description
Upload Results	STRUCTURE output files are provided to Structure Harvester.
Visualization	Generates line graphs of likelihood values for each K.
ΔK Calculation	Applies Evanno’s method to detect the optimal K value.
Output	Provides plots and tables summarizing genetic clustering.

Using STRUCTURESelector

STRUCTURESelector is an advanced online platform that integrates multiple statistical approaches beyond the Evanno method. It supports Puechmaille’s methods, MedMedK, MedMeanK, and others, providing more robust inference when ΔK alone is insufficient.

STRUCTURESelector accepts results directly from STRUCTURE or Structure Harvester, producing interactive plots, summary statistics, and downloadable reports.

Applications in Plant Breeding

In plant breeding and crop genetics, understanding population structure is vital for:

• Identifying distinct genetic groups for association mapping
• Detecting admixture and hybridization
• Improving the accuracy of GWAS (Genome-Wide Association Studies)
• Designing effective breeding strategies based on genetic diversity

By combining STRUCTURE, Structure Harvester, and STRUCTURESelector, breeders gain deeper insights into the genetic architecture of germplasm collections.

Conclusion

Structure Harvester and STRUCTURESelector extend the functionality of STRUCTURE by providing clear, reliable methods to interpret genetic clustering. These tools are indispensable for modern plant breeding, helping researchers make informed decisions about germplasm selection, population analysis, and breeding program design.