GENOME-WIDE ASSOCIATION STUDY FOR AGRO-MORPHOLOGICAL TRAITS AND PIPELINE DEVELOPMENT TO ASSESS POPULATION STRUCTURE IN FINGER MILLET

Abstract

Finger millet is a small-grained millet that is ranked as the third most important after pearl millet and foxtail millet. There are diverse finger millet varieties that are grown in eastern Africa characterized by diverse agro-morphological traits and genetic composition however not much research has been done on the crop due to limited genetic and genomic resources. Hence, there is a need to develop genetic and genomic resources to help understand the underlying variation in the finger millet varieties. This study provided insight into the phenotypic and genetic diversity, population structure and genomic loci associated with the most important agronomic traits in a panel of 99 diverse finger millet from countries within the eastern Africa region. Field trials were performed at two locations, Pwani University and Maseno University, and the data collected using guidelines provided by the International Board of Plant Genetic Research (IBPGR) was used as phenotypic data. Assessment of phenotypic diversity was done using correlation and principal component analysis. Heritability was also determined to understand the proportion of each phenotypic trait that was due to genetic factors. Phylogenetic inference, genetic diversity, and population structure were done using a single nucleotide polymorphism (SNP) dataset consisting of 9983 SNPs. A genome-wide association study (GWAS) was done using both phenotypic and SNP datasets. A nextflow pipeline for population structure and genetic diversity was developed to provide a conducive computing environment for running the software intensive analyses. Phenotypic diversity was observed for the accessions from the 11 countries used in both field trials. Cluster analysis, using principal component analysis, a maximum likelihood phylogenetic tree and STRUCTURE identified three major populations within the accessions used. A total of 15 significant SNPs were identified that were associated with the various phenotypic traits, six with days to 50% flowering, two with days to 50% maturity, four with plant height and three with finger number. Candidate genes, responsible for the traits studied, were identified through mapping the SNPs to the finger millet genome. The functions of the genes were identified through comparative analysis to crops like maize, sorghum, rice, pearl millet and Arabidopsis. Genes encoding F-box proteins, disulfide-isomerase, and kinase-like proteins that are responsible for various processes like signal transduction, pollen development, floral transition, seed development tissues and regulation of apoptosis in endothelial cells among others were identified. Overall, the study provides insights into the genetic diversity and population structure of the diverse finger millet varieties studied and identified the genomic regions that control these traits. These findings will inform the development of genetic and genomic resources for finger millet which will aid molecular breeding programs.