Identification of putative drought-responsive genes in rice using gene co-expression analysis

Abstract

Drought is one of the major abiotic stresses causing yield losses and restricted growing area for several major crops. Rice being a major staple food crop and sensitive to water-deficit conditions bears heavy yield losses due to drought stress. To breed drought tolerant rice cultivars, it is of interest to understand the mechanisms of drought tolerance. In this regard, large amount of publicly available transcriptome datasets could be utilized. In this study, we used different transcriptome datasets obtained under drought stress in comparison to normal conditions (control) to identify novel drought responsive genes and their underlying molecular mechanisms. We found 517 core drought responsive differentially expressed genes (DEGs) and different modules using gene co-expression analysis to specifically predict their biological roles in drought tolerance. Gene ontology and KEGG analyses showed key biological processes and metabolic pathways involved in drought tolerance. Further, network analysis pinpointed important hub DEGs and major transcription factors regulating the expression of drought responsive genes in each module. These identified novel DEGs and transcription factors could be functionally characterized using systems biology approaches, which can significantly enhance our knowledge about the molecular mechanisms of drought tolerance in rice. Background: Rice is a model plant species and a major staple food crop among cereal feeding around half of the world population [1]. Since, rice has very high-water requirements for good production, drought is a major cause for its yield reduction and limits the growing area worldwide [2]. Several projects were launched for the annotation of rice genome which enabled researchers to utilize the genome-annotation information for the functional characterization of rice genes. One of the promising projects is Rice Genome Annotation Project by MSU (http://rice.plantbiology.msu.edu/index.shtml) with its 7th release providing annotation of all the 12 chromosomes and about 40,000-60,000 estimated rice genes [3]. Recently the reduced cost of transcriptome analysis and availability of efficient omics tools have shifted the attention of molecular biologists to utilize the genome annotation information via computational approaches. Since most of the biological pathways involve