Discovering common and population-specific QTLs for leaf rust resistance in different Barley populations.

key messageMulti-population GWAS lead to identification of common and population-specific QTLs for leaf rust resistance in barley. Genome-wide association studies (GWAS) are a powerful tool for detecting genetic markers associated with traits of interest. However, these studies are typically restricted to a single population, and transferability of identified marker effects across populations is challenged by population differences in linkage, allele frequencies, epistatic effects, and environmental context. When comparing GWAS results between populations, a lack of overlapping signals is often interpreted as a lack of common quantitative trait loci (QTLs), although such discrepancies may result from differences in statistical power to detect signals. In barley (Hordeum vulgare L.), where genetic leaf rust resistance is rapidly overcome by evolving pathogens, identification of cross-population robust and potentially transferable resistance loci is a key task. Here, we present a mixed model approach for multi-population GWAS that estimates correlated marker effects in multiple populations and use this to test for significant effects across and within populations. Applying this model to four barley breeding populations revealed both common and population-specific QTL effects for leaf rust resistance, including loci colocalizing with known Rph genes and novel regions with plausible candidate genes. Multi-population GWAS increased power, revealing signals not detected by GWAS within populations. We categorized the reported QTLs into three groups based on marker-associated allele effects: (1) consistent effect direction across populations, (2) differing effect direction across populations, and (3) present in a single population. The study highlights the transferability and limitations of leaf rust resistance QTLs across different barley populations and provides a general statistical framework to support robust marker-assisted selection across populations.