Agrematch announces trials success for AI-discovered natural herbicide

Agrematch Ltd, an AI company focusing on the discovery and development of crop input products, announced that it has moved closer to commercializing its novel bioherbicide after favourable trials with commercially important crop weeds, including strains resistant to widely used synthetic herbicides.

Discovered and developed in less than 12 months using Agrematch’s artificial intelligence predictive system and validation platform, AI4AI (Artificial Intelligence for Active Ingredients), the new active ingredient – codenamed AS9057 – displayed superior post-emergence control of various weed species, including those of the Amaranthus and Chenopodium genera.

AS9057 is also believed to have a new mode of action, making its discovery highly significant for global crop production and the $26bn global herbicides market. Since the 1980s, just two new herbicide modes of action have been commercially introduced; over the same period, incidence of resistance has soared by more than 600%.

Initial assessment of AS9057 indicates its suitability both as a standalone product and in mixtures with existing synthetic active ingredients. This supports Agrematch’s objective to lead the transition from synthetic chemistries to natural solutions, helping to reduce overall volumes of synthetic pesticides and allowing farmers to combine herbicide modes of action – a crucial step in resistance management.

While withdrawal of outdated and harmful products helps meet consumer demand for food grown in a more sustainable and environmentally friendly manner, it leaves farmers with limited solutions to manage resistance. Agrematch aims to use its AI discovery platform to tackle this resistance problem by satisfying growers’ urgent need for safe, sustainable and effective new products, potentially with undiscovered modes of action.

The AI approach also ‘fast-tracks’ the discovery of new active ingredients. Traditional crop protection R&D can take up to four years to manually screen hundreds of thousands of chemical compounds in a laboratory. This is typically followed by up to six years of product development and regulatory processes, requiring an average investment of $285m.

In contrast, AI4AI promises to bring new active ingredients to the market in as little as five years, reducing the cost to between $10m and $60m and hugely reducing the downstream risk in crop protection development.

AI4AI is described by Agrematch as a ‘browser to explore the vast space of compounds and their functions’. The system applies unique machine-learning and deep-learning capabilities to a large proprietary database.

Capable of bringing into play chemical spaces that are inaccessible by traditional methods, AI4AI can predict key attributes of the compounds it discovers, allowing researchers to assess bioactivity, mode of action, toxicity profile and COGS, critical factors affecting the compound’s chance of becoming a commercial product at a much earlier stage.

Eitan Glick, CEO of Agrematch, stated: ″Our AI predictive technology is revolutionizing the discovery of novel, safe and effective bioactive compounds needed to grow food sustainably at scale.

″The discovery and progress of our first bioherbicide is another proof point for the AI4AI computational predicting platform and its ability to accelerate product development.

″We’re looking forward to this bioherbicide progressing in the development pipeline and becoming an important tool for farmers to control weeds.″

Ifat Shub, CTO of Agrematch, added: ″Our vision is to lead a transition in the crop protection industry, from synthetic to natural products, allowing growers to address their challenges sustainably while meeting consumers’ demands.

″AS9057 is an important milestone in Agrematch’s journey to realising its vision and we’re excited to see it progress.″