Aberystwyth Team Targets Wild Traits With 43,000-Seed Bank and AI

A seed bank holding more than 43,000 samples from around the world sits at the centre of a three-year Aberystwyth University project that aims to move useful wild-plant traits into commercial crops through conventional breeding.

Researchers at the university's Institute of Biological, Environmental and Rural Sciences (IBERS) will investigate how chromosomes recognise one another during reproduction. Biological barriers in distantly related plants routinely block those pairings, which keeps drought, disease and stress tolerance locked in wild relatives rather than in farmers' fields. The Biotechnology and Biological Sciences Research Council (BBSRC), part of UK Research and Innovation, funds the work.

"What attracts one thing to another is one of the great mysteries of life, whether in plants, animals or even humans," said project leader Dr Andrew Lloyd of IBERS. "This research will not solve all of those mysteries but we hope it will help make crops more resilient to future challenges."

What blocks trait transfer today?

During reproduction, chromosome pairs find each other and exchange sections of DNA, creating new trait combinations in the next generation. In distantly related plants, differences between chromosomes can stop that exchange from working. That failure is the bottleneck breeders face when they try to pull a valuable characteristic from a wild relative into an established crop line.

The project will combine Aberystwyth's plant breeding expertise, the IBERS seed bank's 43,000-plus samples, advances in DNA sequencing and chromosome painting to examine chromosome interactions directly.

"Until recently we have not had the tools needed to answer these questions," Lloyd said. "This project will take advantage of cutting-edge DNA sequencing and chromosome painting tools recently developed by ourselves and others to look at the problem in completely new ways."

Where does AI fit in?

The team pairs IBERS biologists with the university's Department of Computer Science. Dr Faisal Rezwan, senior lecturer in bioinformatics, said computing advances now make previously unreachable questions answerable.

"Only now do we have the computational and experimental tools required to decode the complex interplay between how chromosomes differ from one another and how they interact," Rezwan said. "By integrating innovations in both computing and biological sciences, this project will address fundamental questions that have previously remained beyond our reach."

Lloyd frames the seed bank as an underused asset. "I'm excited about the potential this project holds for unlocking the incredible diversity of useful traits currently held in our seed banks. These collections are a vast and valuable resource," he said. "If we can use them to understand the rules governing chromosome recognition, we open the door to harnessing those traits to help crops withstand the pressures of a changing climate."

On AI's role, he pointed to prediction rather than description. "We can use rapidly developing AI tools to understand and predict outcomes in ways that simply were not possible before," Lloyd said. "My hope is that this will reveal the most efficient routes for transferring valuable traits between distantly related plants, accelerating the development of the resilient crops we'll need in the years ahead."

What does this change for growers? Nothing yet in a catalogue. But if chromosome recognition rules prove transferable, breeders gain a cheaper, conventional route to wild-relative traits — and a three-year clock, funded by BBSRC, is now running toward that answer.