QUANTITATIVE GENETICS APPROACHES FOR PREDICTING HYBRID VIGOR IN AGRICULTURAL BREEDING PROGRAMS

Abstract

Background: Hybrid vigor or heterosis is a fundamental phenomenon in agricultural breeding which results in higher crop productivity, growth performance and stress tolerance in hybrid populations. Accurate prediction of hybrid performance is a great challenge for modern breeding programs due to the complexity of genetic interactions and environmental effects.

Objective: The aim of this study was to evaluate quantitative genetics approaches for predicting hybrid vigor with genomic selection, combining ability analysis, and machine learning-assisted breeding strategies.

Methods: The maize, rice and wheat breeding populations were employed for the collection of high-density SNP marker datasets, phenotypic trait measurements and environmental data. Genetic effects (additive, dominance and epistatic) related to hybrid performance were analyzed with statistical models including genomic best linear unbiased prediction (GBLUP), general combining ability (GCA) and machine learning algorithms.

Findings: Genomic prediction models obtained prediction accuracies >90% for yield-related traits, and machine learning-assisted genomic selection improved hybrid classification efficiency by about 25%. The combining ability analysis successfully identified the superior parental lines associated with improved biomass accumulation and stress tolerance under varying environmental conditions.

Conclusion: The results show that the combined use of quantitative genetics and genomic prediction methods can significantly enhance the efficiency of hybrid breeding, help accelerate crop improvement programs and promote sustainable agricultural productivity under changing climatic conditions.