Identification of core drought-responsive genes for developing drought-tolerant rice varieties through meta-analysis of RNA-Seq data

Rice production is significantly affected by drought at various developmental stages. The identification and shortlisting of candidate genes for developing drought-tolerant rice varieties are ongoing. To identify core drought-responsive genes, we compared the transcriptomic changes between ten drought-tolerant rice varieties. Through our meta-analysis of publicly available RNA-Seq datasets, we identified 3080 core differentially expressed genes (DEGs) in response to drought stress. K-means clustering analysis showed that the expression of core DEGs can be grouped into two clusters with distinct expression dynamics under drought stress. Gene ontology enrichment analysis revealed that the functional distribution of 1282 genes in cluster 1 showing upregulation in all the tolerant lines was mainly related to the abiotic stress response. A weighted gene co-expression network analysis of all the genes in cluster 1 revealed that among the 11 co-expressed modules, the blue module showed significant positive correlation (195 genes; R-2 = 0.73, P = 3.0 x 10(-19)) with drought stress, which assisted us to narrow down the genes to identify candidates for drought stress tolerance. Through a literature search to identify functionally characterized genes, six drought tolerance-related candidate genes were detected, including LEA4-5 (LOC_Os08g23870), LEA14 (LOC_Os01g12580), OsRab16A (LOC_Os11g26790), OsADF3 (LOC_Os03g60580), OsPP108 (LOC_ Os09g15670) and GBF3 (LOC_Os01g46970). Finally, we constructed a protein- protein interaction network and selected the hub genes in the blue module, two of which overlapped with candidates and may play a critical role in drought tolerance. Our results provide promising candidate genes that may be useful in transgenic studies or molecular marker-assisted selection for improving drought tolerance and increasing rice yields.