Genomic Selection Simulation for Breeding Cycle Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional plant breeding methods face limitations due to environmental noise and the difficulty in identifying all quantitative trait loci (QTLs) associated with desirable traits, leading to reduced genetic gain and inefficiencies in marker-assisted selection (MAS) and genome-wide selection (GWS).
Innovation Solution
The method involves simulating crosses between breeding pairs to predict progeny with increased genetic gain by calculating genetic potential values based on genome-wide markers, selecting optimal breeding pairs, and repeating the process to achieve desired phenotypes or genotypes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If phenotypic selection is used to select superior progeny, then selection can be performed based on observable traits, but environmental noise reduces selection efficiency and limits breeding cycles to one per year
Solution Approach 1:
The patent replaces phenotypic selection (mechanical/physical observation) with genomic selection using molecular markers. Instead of observing physical traits that are influenced by environment, the invention uses DNA markers to predict breeding values, eliminating environmental noise and enabling multiple selection cycles per year through in silico evaluation.
Solution Approach 2:
The patent introduces molecular markers as an intermediary between genotype and phenotype. These markers serve as proxies for QTLs and enable indirect selection of desirable traits without being affected by environmental conditions, thereby improving selection accuracy and enabling faster breeding cycles.
2Productivity
If marker-assisted selection (MAS) is used to speed up breeding, then multiple cycles per year are possible, but difficulty in identifying all QTLs reduces overall effectiveness
Solution Approach 1:
The patent applies genome-wide markers that can simultaneously detect all QTLs across the genome, not just pre-identified ones. This universal approach allows the same marker set to serve multiple functions: detecting known QTLs, discovering new QTLs, and capturing small-effect QTLs that MAS would miss, thereby preventing information loss.
Solution Approach 2:
The patent uses an excessive number of markers covering the entire genome rather than a limited set of known QTLs. This ensures that even small-effect QTLs and previously unidentified genetic factors are captured, preventing information loss while enabling rapid multi-cycle breeding.
3Ease of operation
If QTL effects are estimated in MAS, then targeted selection can be performed, but QTL effects are often overestimated which reduces selection efficiency
Solution Approach 1:
The patent replaces QTL effect estimation (prone to overestimation) with genomic breeding value prediction using all available markers. This substitution uses a more robust statistical approach that accounts for the combined small effects of many markers, eliminating the overestimation problem while maintaining targeted selection capability through predicted GBVs.
Solution Approach 2:
The patent merges the effects of all genome-wide markers into a single genomic breeding value prediction, rather than estimating individual QTL effects. This combination approach prevents overestimation by distributing the genetic value across many markers, while still enabling targeted selection of superior individuals based on their predicted GBVs.
Data Source
AI summary
Methods for method for increasing genetic gain in a breeding process are provided. Also provided are methods for choosing breeding pairs predicted to produce progeny having desired phenotypes, methods for increasing the likelihood of producing progeny individual having desired phenotypes, methods for generating progeny individual having desired genotypes and/or phenotypes, progeny produced thereby, and cells, seeds, parts, and tissues cultures thereof.