Collaborative Research: TRTech-PGR: Functional Pangenomic and Machine Learning Tools for Exploring Genetic Diversity in Plants

As climate change accelerates, the ability of plants to adapt will be critical in ensuring their ability to survive in future decades. A plant population that has a more diverse set of genes and other genetic material is more likely to survive climate shifts and climate extremes. However, as plants are forced to adapt to harsher conditions, many plants will die off, reducing the diversity of the genes available for future adaptation of the population.

Because changing climates are requiring organisms to adapt at much faster rates than in the past, understanding the genetic diversity present in a population and how that diversity is changing is critical for understanding which plants are in danger. This knowledge is critical to maintaining strong food systems and ecosystems. Breakthroughs in DNA sequencing technology and artificial intelligence enable, for the first time, the ability to look at the genes and genomes in many individuals of the same plant species at once to understand their genetic diversity.

This project will use this computational approach to connect different versions of genes with successful growth in different climate conditions, revealing which versions of specific genes plants will need as the climate around them changes. The methods developed will be incorporated into a software tool that will help scientists study genetic diversity in many different plant species and other organisms.

Additionally, our K-12, undergraduate, and graduate outreach programs will help train scientists and enable them to use the software tool to combat the effects of climate change.

Given rapid and unprecedented changes in climate that threaten global food security and increase plant extinction risk, the need for methods that enhance and expedite insights into plant genetic diversity critical for adaptation is widespread and immediate. The exponential growth of complete, high-quality genomes, functional sequencing data sets that help delineate promoters and enhancers, and large-scale phenotypic data sets, has created opportunity for algorithmic innovation to better understand extant genetic diversity, how it is changing, and the genes and regulatory sequences that drive adaptation.

This project will create a modular software tool for biologists that will enable identification of cis-regulatory elements using functional sequence and pangenome-wide nucleotide signals, connecting these and other genomic features to expression data and climate variables. This tool will combine a novel graph model with recent innovations in artificial intelligence, improving pangenomic analyses by incorporating annotation of genes and their control regions with clade-wide genomic variation and phenotypic variance to identify genomic drivers of phenotypic differences that enable adaptation.

To test and characterize the tool, this project will focus on publicly available data sets from plants, whose complex genomes provide an important stress test and whose genetic diversity is critical to agricultural systems, food security, and ecosystem health. The proposed tool has cross-disciplinary implications, allowing any researcher with genome assemblies, functional sequence, and phenotypic data to perform unbiased, comprehensive analyses of a clade’s genetic diversity and explore how that diversity drives phenotypic adaptation.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.