Bionote
Dr. Jenifer Camila Godoy holds a Ph.D. in Genetics and Plant Breeding from the University of São Paulo and is currently a Postdoctoral Research Associate in the Lipka Lab at the University of Illinois Urbana Champaign. Her research focuses on developing innovative approaches to integrate gene co-expression networks into genomic selection models, aiming to improve the accuracy of predicting maize hybrid performance in non-evaluated environments.
Presentation Abstract
Climate change, land degradation, population growth, pandemics, and conflicts are making the global food system increasingly vulnerable. Advanced plant breeding technologies like genomic selection (GS) can help address these challenges by enhancing agricultural productivity and ensuring food security. GS uses thousands of molecular markers to predict the genomic estimated breeding values (GEBVs) of individuals not yet field-tested, enabling shorter breeding cycles, reduced phenotyping costs, and increased genetic gains, mainly in situations where traits pose difficulties in terms of measurement or cost or when various environments need to be assessed. In maize breeding, GS is especially useful for predicting the performance of new hybrids, given the impracticality of evaluating all possible crosses. However, conventional GS models based solely on SNP markers may not fully capture the molecular complexity of trait expression under environmental variation. To overcome this limitation, transcriptomic data, such as gene expression levels and co-expression patterns, can be integrated to improve prediction accuracy. This research aims to evaluate the potential of incorporating transcriptomic data into GS models. Specifically, the objectives are to: (i) quantify gene expression variation in elite inbred lines and hybrids under different environmental conditions; (ii) construct co-expression networks to identify modules responsive to environmental variation; and (iii) integrate these transcriptional features into GS models to assess their predictive ability. Maize seedlings will be profiled under control, heat, and cold conditions. Transcriptomic variation will be analyzed and incorporated into statistical models to predict agronomic traits across multiple environments. Model performance will be evaluated via cross-validation and compared to conventional SNP-based models. Additionally, stochastic simulations will explore practical strategies for using gene expression in breeding pipelines. This work is expected to generate insights into the molecular basis of trait variation and offer guidance for improving prediction accuracy and the development of climate-resilient maize varieties.
Co-Researcher (Advisors): Alexander Lipka, Martin Bohn