- A crop’s genetic background can determine whether a gene edit succeeds.
- As gene editing scales up, breeders are increasingly focused on genetic context.
- AI is helping breeders predict how gene-edited traits will perform in the field.
For years, biotechnology has focused on one central challenge in agriculture: making precise changes to crop DNA. Advances in CRISPR and other gene-editing technologies have brought scientists closer than ever to that goal.
Yet as gene editing moves from the laboratory into commercial crop development, researchers are increasingly grappling with a different question: what happens in the field after an edit is made?
The challenge now is understanding how that edited gene behaves within the broader genetic makeup of a crop, said Dr Thomas Brutnell, CEO and co-founder of DNA Harvest Partners, an AI-guided predictive breeding company headquartered in Singapore.
“Gene editing opens the door to precise, targeted improvements in crops — but the genetic background in which an edit is expressed can make or break its performance in the field,” Brutnell explained.
Genetic background refers to the wider set of genes already present in a crop variety, which can influence whether a new edit performs as expected.
An overlooked issue
Speaking to AgNavigator, he said the industry has paid very little attention to this issue, partly because gene-edited crops — such as the Sanatech GABA tomato — remain relatively scarce.
“However, as major seed companies begin stacking multiple edited alleles into row crops, there will be a need for both high penetrance — that is, a strong effect across a diversity of lines — and favourable interactions with other edited alleles.”
Dan Dong Yul Sung, executive vice president of ToolGen’s Seed Research and Business Division, believes the industry broadly understands the importance of genetic context in commercial variety development.
However, the concept has not yet been fully exploited because researchers still lack a comprehensive understanding of how different genetic backgrounds influence crop traits – something he believes is changing with AI.
“With the advent of AI and as information on how genetic context influences crop traits accumulates, the industry will utilise more of the genetic context,” said Sung.
Brutnell believes this issue is particularly relevant to gene editing compared to genetic modification (GMO) because edited genes remain embedded within the plant’s existing biological networks.
Sung echoed this view.
“For gene editing, it matters more because you are improving endogenous genes that are constantly interacting with other genes in the genome, whereas foreign DNA in GMO products typically has less interaction with the host genome and consequently can work without much influence from the genome.”
Sung highlighted naturally occurring mutations in the ALS gene, where similar changes can produce different levels of herbicide sensitivity across weed species, as an example of how genetic background can affect trait outcomes.
A strategic partnership
Against this backdrop, DNA Harvest Partners and ToolGen have entered into a strategic partnership aimed at accelerating the development of improved rapeseed (Brassica napus) varieties.
The partnership will combine DNA Harvest Partners’ artificial intelligence-driven breeding technologies with ToolGen’s CRISPR-Cas9 gene-editing capabilities.
According to DNA Harvest Partners, its SmartCross platform can simulate millions of potential breeding crosses within ToolGen’s rapeseed germplasm to predict traits such as yield potential, seed oil quantity and seed oil quality before extensive field trials begin.
Its GEO platform is then used to identify genetic combinations that may help an edited trait perform more consistently.
“By combining ToolGen’s CRISPR-Cas9 technology with DNAHP’s AI-guided breeding platform, we aim to develop better crop varieties faster, reduce development time and cost, improve the probability of field success, and deliver differentiated crop varieties that address the practical needs of farmers, seed partners, and the broader agricultural value chain,” said Dr Jongsang Ryu, CEO of the South Korean company.
The combined use of predictive breeding and gene editing could produce new varieties at “half the cost and half the time of traditional breeding”, said Brutnell.
“The goal is to cut the cost and development times of generating new varieties by 50% or more. In doing so, consumer traits, rather than grower traits, can take centre stage,” he said.
Brutnell cited emerging products such as seedless blackberries, non-browning bananas and the GABA tomato as examples of consumer-facing traits that could become more commercially viable if crop development becomes cheaper and faster.
Future plans
Beyond the rapeseed programme, Brutnell sees broader opportunities for the partnership model.
“Although we are now focusing on ToolGen’s germplasm collection, we are also excited about extending this partnership to additional third-party seed companies that are looking to accelerate both their adoption of gene-editing technology and their use of the cost savings and speed of our predictive breeding platform to develop new traited varieties at half the cost and half the time of traditional breeding.”
Both companies expect AI to play a growing role in future breeding programmes, with Brutnell highlighting how AI-guided breeding is already becoming common among major seed companies.
“We do and this is already happening with the major seed companies. However, for hundreds of seed companies across Asia-Pacific, this technology is out of reach. Extending this collaboration to third-party seed companies would be one way for those companies to gain access to the industry’s most advanced breeding and gene-editing technologies.”
