A London-based biotech company is growing the 99% of soil microbes that have never been cultured in a lab, then screening them for new fungicides to fight a wheat-killing fungus. Wheat is the UK’s main arable crop, worth £1.2 billion annually, but the fungus *Zymoseptoria tritici* can destroy half the yield. Farmers currently spray up to 12 times per season with synthetic fungicides, but the pathogen is evolving resistance and the chemicals carry environmental and health risks. Existing microbial fungicides come from only 1% of microbes—those that survive isolation. Bactobio’s platform uses next-generation sequencing and machine learning to recreate the natural conditions that keep the other 99% alive, giving them access to an untapped chemical arsenal. If successful, the project will deliver 3–5 new bio-derived fungicides tested up to small-plot field trials. These compounds could replace or reduce synthetic sprays, slowing resistance, cutting environmental damage, and protecting the UK wheat supply chain. The company plans to license the fungicides to large agritech firms for commercial scale-up.
View original technical description
Modern farming relies on synthetic fungicides to support crop yields. Unfortunately, these chemicals are often associated with environmental and human health risks and are increasingly failing due to development of resistance. In the UK, wheat represents the primary arable grain crop, with an annual production of 15.5Mn tonnes and a market value of £1.2Bn. _Zymoseptoria tritici_, a fungal pathogen, causes the most destructive and economically damaging foliar disease affecting winter wheat in the UK, resulting in yield losses of **=**50%. To manage this disease, farmers resort to **=**12 fungicide applications, but this widespread use is causing pathogen resistance and environmental degradation. There is therefore an urgent need for new, safe, and green fungicides and microbial chemistry offers a powerful alternative. Microbes that compete with pathogens in nature provide a source of highly effective fungicidal compounds, but historically only 1% of these microbes have been explored. This is because only this small percentage can survive the transition from natural symbiotic communities to isolation in laboratory conditions. At Bactobio, we use breakthroughs in next-generation sequencing, machine learning, and bioengineering to understand microbial community compositions and chemical/physical conditions needed to make each microbial community viable. We mimic these communities and conditions, resulting in novel cultured microbial species, which we screen for novel, green, and safe fungicides. We have developed a platform that provides access to the remaining 99% of previously unexplored microbes. The extensive database we are generating supports machine learning approaches to further optimise our experimentation, ensuring that we harness the full potential of uncultured microbes. In this project, our goal is to leverage our exclusive access to previously uncultured microbes to discover new bio-derived fungicides against _Z. tritici_. We aim to enhance our microbial library and build a sophisticated compound identification platform, enabling the development of a high-throughput fungicide discovery engine. We intend to boost our genomics and compound profiling technologies to consistently discover, test, and scale up the production of new compounds for further development. Our objective is to discover and develop fungicidal compounds targeting _Z. tritici._ By project end, we aim to have 3-5 fungicides targeting _Z. tritici_ with complete data packages up to small plot field trials. Subsequently, we will license them to large agritech players. Together, we aim to safeguard future food security, reduce the environmental impact of disease management, slow the spread of disease resistance, and protect the economic viability of wheat farming in the UK.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know