Finger millet is a grain crop of strategic importance to food security in Eastern Africa. The grain has high nutritional value, can grow in arid environments and thus is important to the livelihood of smallholder farmers. A major agricultural goal in the region is to develop higher yielding varieties of finger millet through reducing or eliminating diseases that impact growth of the plant. Blast fungus is a pathogen that reduces yield up to 80% and is one of the main diseases affecting finger millet. To understand how to control disease outbreaks, this project uses genomic sequencing as a powerful approach to identify precise strains of the fungus and to study how the fungus causes disease symptoms in the plant. Sequence analyses of blast strains collected in Kenya, Tanzania, Uganda and Ethiopia will provide information on the genetic diversity of the pathogen in Eastern Africa, and provide a resource to identify the factors that are responsible for infection of finger millet. The knowledge from this approach is essential to develop efficient disease management strategies. Furthermore, sequence analyses of the finger millet host will clarify why some cultivars are more resistant to blast than others. The generated resources will also be used as a vehicle to train undergraduate and graduate students in Eastern Africa in bioinformatics, an expertise that is essential to translate the information to improve breeding strategies.
The specific aims of the project are to (1) Generate 80X PacBio sequence for the allotetraploid finger millet genome (1C=1.8 Gb) to generate a high quality genome assembly (1C=1.8 Gb); (2) Resequence 200 Eastern African isolates of the finger millet blast fungus Magnaporthe oryzae, including 24 that were collected 10 years ago, to determine the diversity and evolution of this finger millet pathogen both over time and across geographic regions. The blast genome sequences will be mined to identify candidate effector genes using an effector prediction pipeline that incorporates common characteristics of known effectors (secretion and high polymorphism levels;(3) Analyze the blast-finger millet interaction transcriptome using RNA-Seq to identify genes that are induced at early stages of infection. Genes encoding secreted proteins will be identified from the RNA-Seq experiment and cross-referenced to those identified using the effector prediction pipeline. Host genes that are differentially expressed will be compared between compatible and incompatible interactions, and with genes that are differentially expressed during early stages of blast infection in rice, and(4) Develop a nested association mapping panel of some 4000 RILs derived from 21 diverse parents using a double round robin design. This population will represent the first mapping resource that captures substantial diversity present in finger millet germplasm and has a high quantitative trait loci detection power.