*2.4. Identification and Validation of DEGs*

To adapt to stresses such as drought and high temperature, plants usually actively regulate the expression of endogenous genes. To investigate the dynamic transcriptional profiles in response to heat in anthers of SDWG005, changes in the expression of gene transcripts during heat treatment (including WD\_1, WD\_2, WD\_6, WD\_12, and WD\_0 as a control) were profiled at the genome-wide level by RNA-seq. The expression level of each transcript/gene under each treatment (e.g., WD\_1) was identified by the mean value of three biological replicates (e.g., WD\_1\_1, WD\_1\_2 and WD\_1\_3). All of the correlation coefficients between the three biological replicates for each treatment based on all of the transcripts were more than 0.9, indicating that the expression data were highly reproducible (Table S1).

Genes that showed at least a two-fold change in expression in the heat-treated samples compared to the control sample were referred to as DEGs in this study. In total, 3559 genes were identified as DEGs at a minimum of one time-point. Based on the 3559 DEGs, three biological replicates for each sample could be clustered together by hierarchical cluster analysis (Figure S1), which indicated that the DEGs identified were reliable. The number of upregulated and downregulated DEGs varied considerably in different treatments. An increase in the number of DEGs was clearly observed as the heat treatment period lengthened, with 477, 869, 2335, and 2210 DEGs identified after 1 (WD\_1), 2 (WD\_2), 6 (WD\_6), and 12 (WD\_12) h of heat treatment, respectively (Figure 4a). All four heat

treatments produced more downregulated genes than upregulated genes, especially in the short-term heat treatments (Figure 4b).

**Figure 4.** Analysis of differentially expressed genes in response to heat stress. (**a**) Venn diagram analysis of significantly differentially expressed genes at the four heat stress time points. (**b**) Numbers of significantly regulated genes at the four heat stress time points. Red and green columns represent the number of up- and down-regulated genes at different heat stress time points, respectively.

To further validate the RNA-seq results of heat stress responsive genes in this study, qRT-PCR was conducted on anthers of SDWG005 under the same heat treatments as for RNA-seq to examine expressions of ten randomly selected DEGs. The expression patterns of all ten genes examined by qRT-PCR coincided with those determined by RNA-seq analysis (Figure 5), also indicating that the RNA-seq analysis was reliable. For example, RNA-seq analysis revealed 2–4.5 times higher expression levels of LOC\_Os04g01740 (*OsHSP1*) in the heat shock (HS) samples than in the control, and the qRT-PCR analysis revealed 1.5–4 times higher expression levels of this gene relative to the control.

**Figure 5.** Expression levels of ten randomly selected DEGs by RNA-seq and qRT-PCR. RNA-seq: expression change of these genes in anthers of SDWG005 under heat treatments as compared to control by RNA-seq. WD-qPCR: expression change of these genes in anthers of SDWG005 under heat treatments as compared to control by real-time PCR.
