Topics Map > Services > Research Computing and Support > CCAST
Highlights - Increasing Pulse Crop Productivity Using Genomics, High-Throughput Phenotyping, and Modeling
Principal Investigator: Nonoy Bandillo (Plant Sciences, North Dakota State University)
Dr. Bandillo’s Lab at NDSU releases new cultivar of pea, chickpea, and lentil for the Northern Great Plains region. Current research interests include projects geared towards the more basic side of pulse genetics research, such as identification and study of individual genes for important diseases (Aphanomyces and Fusarium root rot, Fusarium wilt, and Ascochyta blight), seed quality and emerging abiotic stresses (salinity and waterlogging). They also explore different tools that aim to increase genetic gain in pulse crops, including genomics-assisted breeding, speed breeding through rapid generation advancement, and unmanned aerial systems for high-throughput phenotyping.
Figure 1: Development of resources at different level of biological organization in pulse crops. Figure adopted from Kremling et al., G3 Genes|Genomes|Genetics 9, 3023 (2019).
They have extensive experience in analyzing large amount of data at different levels of biological organization, including DNA-, RNA-, and phenotype-level (Fig. 1), and been using local high-performance computing (HPC) resources provided by CCAST for analyzing the large volume of sequencing, expression, and phenotypic data. For example, the ongoing resequencing experiment requires up to tens of terabytes of generated sequencing data.
Figure 2: An overall scheme to harness genetic diversity through genomic prediction in the USDA Pea Germplasm Collection. This project will develop genomic resource (training and validation sets) and build prediction models for identification of useful diversity. Many questions surrounding the implementation of genomic selection will be answered so it is used most effectively in public breeding programs.
The Lab is also developing quantitative genetic modeling including the evaluation of genomic prediction for applied breeding efforts. A genomic prediction project will construct a training set of entries for testing and optimizing genomic prediction models; see Fig. 2. They argue that a genomic prediction approach targeting complex traits can help to exploit the genetic diversity stored in gene banks. Genomic estimated breeding values from prediction represents genetic merit for those unphenotyped accessions in gene banks and provide plant breeders a valuable information to identify candidate accessions meriting their attention.
Emerging breeding technologies can make a powerful step change for addressing the major bottleneck in germplasm enhancement and speedy development of new cultivars. High-throughput phenotyping could potentially address the issue of collecting biologically meaningful and interpretable genotypes for thousands of new germplasm and breeding lines. Speed breeding is another technology that would permit breeders to turnover generation and reduce the length of the breeding cycle under controlled greenhouse conditions. Genomic selection would allow the prediction of traits for untested breeding lines using only the DNA information.
Their overall goal is to develop improved cultivars and germplasm for the Northern Plains region, and to improve the efficiency of pulse breeding through the development and evaluation of advanced breeding tools and emerging technologies