Abstract
Drought is the major abiotic stress threatening soybean production globally. However, the genetic basis of soybean drought resistance at the germination stage remains largely unknown. In this study, the drought-tolerance phenotypes of a diverse panel of 207 soybean accessions were examined. Leveraging 95,043 high-density single-nucleotide polymorphism (SNP) markers, a total of 58 quantitative trait loci (QTLs) were detected using the restricted two-stage multi-locus genome-wide association study (RTM-GWAS) method, and 10 of these QTLs were considered as large-contribution QTLs that have larger phenotype variation. Furthermore, RNA-sequencing was performed on the roots of two soybean accessions with contrasting drought tolerance. A total of 1,183, 1,354, and 1,581 differentially expressed genes (DEGs) between two soybean accessions after 0h, 12h, and 24h of drought treatment were separately obtained, and 4,012 and 4,586 genes responsive to drought stress were identified at 12h and 24h, respectively. By utilizing these DEGs, a weighted gene co-expression network analysis (WGCNA) was constructed, and 13 distinct modules were obtained, among which four modules were considered as key modules. Subsequently, 40 hub genes were identified in these four modules. In addition, by combining RTM-GWAS and transcriptome analysis, 22 candidate genes underlying large-contribution QTLs were identified. Based on the functional annotations, Glyma.12G141700, Glyma.15G040000, Glyma.05G049300, Glyma.14G105900, and Glyma.15G041100 were regarded as the most possible candidate genes that regulate soybean drought tolerance at the germination stage. The QTLs, key modules, and hub genes discovered in this study will provide a new understanding of the genetic basis of soybean drought resistance at the germination stage and lay a theoretical foundation for the improvement and innovation of high-quality soybean germplasm.