Early stage sparse testing can increase selection accuracy and genetic gain in plant breeding programmes
Atlin, G.N.
Early stage sparse testing can increase selection accuracy and genetic gain in plant breeding programmes - Edinburgh (CIMMYT) : EUCARPIA Biometrics in Plant Breeding Local Organising Committee, 2025. - 1 page
Presented at XIX Conference 2025, EUCARPIA: Biometrics Plant Breeding, 17-19 Sep, Edinburgh, UK.
Low-cost genome profiling allows information from relatives to be used to reduce replication of selection candidates while retaining selection accuracy, creating opportunities to redesign plant breeding pipelines. Early-stage sparse testing uses the genomic relationship matrix (GRM) to sample the target population of environments (TPE) more effectively by distributing related early-stage selection candidates across many testing locations to train a genomic selection (GS) model, rather than concentrating them in replicated trials at a few research stations. Using the GRM, data on related genotypes across farms can be connected and combined to enable genomic prediction across the entire TPE, not just a few sites. A similar redesign is possible in hybrid breeding programmes, where the GRM allows related selection candidates to be evaluated on several testers, permitting general combining ability (GCA) to be estimated early in the selection process. Simulations indicate that, if testing resources are held constant, sparse testing designs can increase selection accuracy both across environments and across testers. Apromising application of sparse testing is to move early-stage testing directly into farmers’ f ields. Most crop production in Sub-Saharan Africa (SSA) occurs on small farms characterized by low input use, multiple biotic and abiotic stresses, and diverse management factors. However, most CGIAR and national breeding programmes conduct early-stage phenotyping at only a few well-managed research stations. Early-stage on-farm sparse testing (OFST) addresses this issue by shifting the evaluation process from research stations to dozens or even hundreds of farms, treating each farm as an incomplete block. This can enhance genetic gain in farmers’ fields by increasing selection accuracy and intensity while shortening the generation interval. Early-stage OFST is currently being piloted in the CIMMYT maize and CIAT bean programmes in SSA.
Text in English
Testing
Genetic gain
Plant breeding
Breeding programmes
Early selection
Early stage sparse testing can increase selection accuracy and genetic gain in plant breeding programmes - Edinburgh (CIMMYT) : EUCARPIA Biometrics in Plant Breeding Local Organising Committee, 2025. - 1 page
Presented at XIX Conference 2025, EUCARPIA: Biometrics Plant Breeding, 17-19 Sep, Edinburgh, UK.
Low-cost genome profiling allows information from relatives to be used to reduce replication of selection candidates while retaining selection accuracy, creating opportunities to redesign plant breeding pipelines. Early-stage sparse testing uses the genomic relationship matrix (GRM) to sample the target population of environments (TPE) more effectively by distributing related early-stage selection candidates across many testing locations to train a genomic selection (GS) model, rather than concentrating them in replicated trials at a few research stations. Using the GRM, data on related genotypes across farms can be connected and combined to enable genomic prediction across the entire TPE, not just a few sites. A similar redesign is possible in hybrid breeding programmes, where the GRM allows related selection candidates to be evaluated on several testers, permitting general combining ability (GCA) to be estimated early in the selection process. Simulations indicate that, if testing resources are held constant, sparse testing designs can increase selection accuracy both across environments and across testers. Apromising application of sparse testing is to move early-stage testing directly into farmers’ f ields. Most crop production in Sub-Saharan Africa (SSA) occurs on small farms characterized by low input use, multiple biotic and abiotic stresses, and diverse management factors. However, most CGIAR and national breeding programmes conduct early-stage phenotyping at only a few well-managed research stations. Early-stage on-farm sparse testing (OFST) addresses this issue by shifting the evaluation process from research stations to dozens or even hundreds of farms, treating each farm as an incomplete block. This can enhance genetic gain in farmers’ fields by increasing selection accuracy and intensity while shortening the generation interval. Early-stage OFST is currently being piloted in the CIMMYT maize and CIAT bean programmes in SSA.
Text in English
Testing
Genetic gain
Plant breeding
Breeding programmes
Early selection