Professor Donal O'Sullivan
Areas of interest
Incremental year-on-year gain in yield obtained through plant breeding represents a major pillar on which global food security rests. It is widely accepted that in order to deliver continued yield increases into the future, plant breeding needs to enhance conventional phenotypic selection with a variety of marker-assisted selection and genetic modification technologies.
My research is focused on implementing methods and technologies for efficient genetic analysis of relevant agronomic and quality traits primarily in two contrasting and complementary crop species: wheat (Triticum aestivum L.) - the most important arable cereal in the UK and faba bean (Vicia faba L.), our most important arable legume crop. Wheat is a polyploid, selfing member of the grass family, while faba bean is a diploid, outcrossing dicot. Nutritionally, combinations of carbohydrate-rich grains and protein-rich pulses form the staple diet of most of the world's population and agronomically, nitrogen-fixing faba bean boosts nitrogen availability and reduces cost and environmental impact when included in a wheat-based arable rotation.
We use high throughput SNP genotyping platforms such as Illumina's iSelect and so-called next generation re-sequencing to sample variation across the genome of interest. With most genotyping activities being outsourced, the major effort of the lab is focused on population development including large panels of inbred lines for linkage disequilibrium (association) mapping, mutant and multi-parent populations, large-scale field and lab-based phenotyping of these populations and quantitative genetic analysis of the very large resulting datasets. Although increasingly, the emphasis of my research is on interactions between multiple loci affecting a trait, between traits, and on the relationship between an integrative trait such as yield and the its many components, specific trait interests include resistance to fungal pathogens of wheat (particularly yellow rust and ergot) and herbicide resistance and nutritional quality in faba bean.
My PhD research, conducted between University College Dublin (Dublin, Ireland) and Université Paris XI (Orsay, France), was on the genome dynamics of Colletotrichum lindemuthianum - the common bean anthracnose pathogen (O'Sullivan et al, 1998). The quest to separate fungal chromosomes using pulsed field gel electrophoresis during my PhD led me into the other major application of the technique - the creation of Bacterial Artificial Chromosome (BAC) libraries in plants and in the course of my postdoc at Long Ashton Research Station, made maize (O'Sullivan et al, 2001) and common bean BAC libraries which went on to be utilised in a variety of physical mapping and positional cloning applications (Gutierrez-Marcos et al, 2004; D'Ovidio et al, 2004). After a brief interlude at the University of Bristol dedicated to the positional cloning of root hair genes, I moved to a project leader position at NIAB, Cambridge in 2003.
Association genetics in plants was just getting off the ground as I moved to Cambridge, and my principal aim was to develop association genetics approaches in crop species. Initially, meaningful numbers of molecular markers were sadly lacking in the key wheat and barley crops and early studies explored haplotype diversity in relation to phenotype across a handful of candidate genes (Chiapparino et al 2006; Cockram et al, 2007). However, over time, ever higher throughput SNP genotyping platforms emerged and we were well placed with large panels of elite varietal material benefitting from extensive historical datasets, and were able to launch successful association genetics projects first in barley (Cockram et al, 2010) and more recently in wheat.
Initially as something of a side activity, I became interested in faba bean, which had been woefully neglected by the academic community in recent decades. Building a genetics toolkit for faba bean has been slow work, but some years on, we have developed panels of inbred lines, a SNP genotyping platform and commenced some phenotypic trait screens (Cottage et al, 2012; Khamassi et al, 2013).