Understanding genetic diversity for Climate-Resilient Crops

Project Leader: Chengdao Li

The world population is expected to double by 2040, and by the end of this century, a 50% increase in agricultural productivity is needed to feed the world. Climate change has increased extreme weather conditions, and Australia is experiencing higher temperatures and lower winter rainfall. To address this, Prof. Chengdao Li, founder of the Western Crop Genetics Alliance (WCGA), a joint research centre between Murdoch University and the Department of Primary Industry and Regional Development (DPIRD), is developing new breeding strategies and climate change-tolerant varieties of major crops.

 
core hours in Setonix
 
TiB of storage in Acacia
 
researchers
 
Genomes analysed
Partner Institution: Murdoch University and WA Department of Primary Industry and Regional Development (DPIRD) System: Acacia, Setonix, Nimbus, Data Portal Areas of science: Genomics and Transcriptomics

The Challenge

It was estimated the world population will double by 2040. By the end of this century, a minimum of 50% increase in agricultural productivity is required to feed the world. This has put tremendous pressure on agriculture, given there is very limited uncultivated land in the world. Climate change has increased the frequency of extreme weather conditions around the world. In recent decades, Australia has seen a shift towards higher temperatures and lower winter rainfall, which has had significant effects on many farmers. Despite these trends, there remains much uncertainty over the long-term effects of climate change on Australian agriculture sector. This necessitates the breeding of new varieties of crops that are highly tolerant to extreme weather conditions and capable of rapid adaptation to climate change.

Traditional breeding is based on long-term artificial selection, which results in the accumulation of beneficial mutations in various landraces and/or cultivars that fit the cultivation requirements. Such approach requires many cycles of selection and breeding to create a new variety with the desired traits. With the development of new Next Generation Sequencing and genome editing technologies, a novel crop breeding strategy has been proposed to de novo develop new varieties. In principle, the de novo breeding strategy starts by selecting the elite foundation materials from wild or existing crop cultivars to meet the new breeding objectives, followed by the rapid introduction of targeted traits by genetic and breeding tools, and ends by creating new crops that harbour beneficial traits compared with current cultivars.

Prof. Chengdao Li, the founder and director of the Western Crop Genetics Alliance (WCGA), a joint research centre between Murdoch University and the Department of Primary Industry and Regional Development (DPIRD), is currently developing new breeding strategies and climate change tolerant varieties of multiple major crops for Australian breeders and growers.

The Solution

The construction of pangenomes for major crops can greatly facilitate breeding research and is a prerequisite for de novo breeding.  WCGA is building the pangenomes of barley, lupin, and oats, the main grains grown in Western Australia. By sequencing the agronomically representative accessions of barley, lupin, and oats using the latest high-throughput Next Generation Sequencing technologies, i.e., PacBio HiFi sequencing, the goal is to construct the complete pangenome of barley, lupin, and oats. The pangenomes are the “new” reference genomic roadmap that contains not only the genome from a single variety but all the genomic characteristics from all the varieties of the species. This allows the research team to characterise structural variations and explore the origins of gene presence and absence variation, which can provide genetic resources to associate genetic features with phenotypes to identify casual variants in varietal performance (e.g., yield, disease, drought, and salt tolerances).

Identifying sequence variations and their associations with different traits provides key information in a consumable form to help accelerate crop breeding practices. This can be achieved in multiple ways. The first approach is to develop a graphical and interactive analytical platform for scientists to effectively explore the pangenome data and perform various analyses. This includes building genome graphs that can capture the full pangenome diversity and using linear approaches where a single reference is used for all comparisons across a population. The information gleaned from this, such as small to large sequence variations, gene copy numbers, etc., can then be associated with changes in phenotypes and traits and further made available to crop breeders through visualisations and genome browsers to accelerate crop breeding practices. The second approach is to integrate the pangenome data with large volumes of phenotypical datasets to develop systematic models for complicated genotype and environmental interactions. New gene loci associated with traits can be identified, and novel haplotypes can be predicted given specific breeding requirements, which will be delivered as an AI breeding and management system for both breeders and growers.

Prof. Li’s team has been using Pawsey’s Nimbus Cloud, to build its own compute cluster and Setonix supercomputer, for highly parallel computing and large memory tasks at WCGA. The construction of the pangenome and the systematic models would not be possible using ordinary computers or servers. For the many different computational and analytical tasks involved in the pangenome projects in barley, oats, and lupin, Pawsey’s storage and processing power are crucial.

The Outcome

The genomic resources from the pangenome projects of barley, oats, and lupin have been developed into online databases with graphical user interfaces that anyone can access. These resources include assemblies, annotations, and variations. A complete suite of visualisation software packages has been developed at WCGA enabling real-time interactive exploratory analyses at the pangenomic level, chromosome level, genic level, and functional level for breeders and scientists. The pangenome genetic resources and visualisation systems will be available to breeders to improve the genetic selection process and deliver improved varieties to growers.

The pangenome enables a systematic understanding of genomic regulation. With phenotypic data collected from large field trials, complex interactions between genomics and the environment can be revealed and modelled, greatly supporting crop breeding and farming practices. Using machine learning and advanced statistical modelling, a prototype of the haplotype prediction method has been developed that can help breeders design novel varieties that can yield optimal performance under specific environmental conditions.

The project will also contribute to training graduate students in both agriculture and informatics science by supporting PhD scholarships and post-docs offerings.

Traditional crop breeding approach takes generations of selection and cross breeding. Our pangenome resources can dramatically reduce the time and costs in breeding new varieties. Using pangenome and AI breeding have the potentials to revolutionise the breeding practice and farming community.
Chengdao Li,
Project Leader.