CR-Elastic Net Genomic Selection for Low-Heritability Breeding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current genomic selection models for animal breeding face challenges with high requirements for the number of SNPs and reference populations, assuming independence among SNPs, which leads to estimation errors and inefficiencies in breeding high-quality varieties, especially for non-family and low-heritability populations.
Innovation Solution
The CR-Elastic Net model is introduced, which evaluates SNP effect sizes by correcting correlation relationships between SNPs, allowing for genomic selection in non-family and low-heritability varieties using a smaller number of individuals, through a repeatable sampling elastic network approach that sets a constant fine-tuning penalty and model cost function, effectively estimating breeding values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional genomic selection models (BLUP or Bayes) are used, then breeding can be conducted with standard SNP panels, but the models assume SNP independence which causes estimation errors and reduces accuracy
Solution Approach 1:
The patent changes the fundamental parameter assumption from SNP independence to SNP correlation. The CR-Elastic Net model explicitly incorporates the correlation structure among SNPs by using a penalized likelihood approach with a correlation-adjusted penalty term, transforming the model's behavior to account for linked SNPs and thereby improving estimation accuracy
Solution Approach 2:
The patent replaces the traditional BLUP or Bayes mechanical framework with a new CR-Elastic Net computational approach. This substitution introduces a correlation-aware penalty mechanism that automatically adjusts for SNP dependencies, replacing the independence-based mechanics of conventional models with correlation-based mechanics
2Measurement precision
If high-density SNP panels and large reference populations are used, then genomic selection accuracy can be improved, but the cost and complexity of breeding programs increase significantly
Solution Approach 1:
The patent changes the penalty parameter structure in the Elastic Net framework to account for SNP correlation. By modifying the penalty term to incorporate correlation information, the model achieves better accuracy with lower-density SNP panels and smaller reference populations, reducing the resource requirements while maintaining or improving selection accuracy
Solution Approach 2:
The patent extracts and utilizes only the essential correlation information among SNPs rather than requiring complete high-density SNP data. By focusing on the correlation structure rather than individual SNP densities, the model achieves effective genomic selection with reduced data requirements, simplifying the breeding program
3Device complexity
If SNP correlation is ignored as assumed in conventional models, then the models are computationally simpler, but estimation errors increase and effective SNPs are diluted by unrelated SNPs
Solution Approach 1:
The patent modifies the penalty parameter in the Elastic Net model to incorporate SNP correlation information. This parameter change allows the model to account for correlation effects computationally efficiently while improving estimation accuracy, avoiding the need for complex iterative procedures that would significantly increase computational burden
Data Source
AI summary
A method and system for genomic selection of a non-family and low-heritability variety are provided. The method includes: evaluating a single nucleotide polymorphisms (SNPs) effect size based on a non-equivalent condition of SNPs by correcting a correlation relationship between the SNPs; and breeding the non-family and low-heritability variety based on the SNPs effect size. In the present disclosure, the method for evaluating the SNPs effect size can be used for breeding the non-family and low-heritability variety.