By applying its advanced AI predictive system, Agrematch can reveal molecules with biological activity, which were previously inaccessible by traditional discovery methods.
Agrematch is disrupting the Ag/Food industries by predicting and developing novel products in a fraction of the time, costs and risks compared to the industry standard.
There is a wide array of applications which require novel bioactive compounds for their products.
Yet the discovery and development of new compounds with specific requirements is very challenging, prohibitively expensive and inherently comprises high risk of late-stage development failure.
Current methods are inefficient and have exhausted the known chemical space. They simply don’t possess the highly advanced tools required to satisfy the market's growing need for sustainable compounds.
More than 1060 compounds are estimated to follow the Lipinski rule and are potentially feasible product candidates.
So far, existing models have had little success in navigating the chemical map and uncovering new meaningful ground.
We have developed a new biochemical language that articulates the needs and lights up spots around novel functional bioactive compounds.
AI4AI: artificial intelligence for active ingredients
AI prediction platform for functional compound characterization and downstream risk mitigation.
We combined the power of our machine/deep learning (ML/DL) algorithms and proprietary database with biology, chemistry, and agricultural know-how to create a powerful integrated AI predictive platform.
In addition to finding desired compounds’ bioactivity, the system is designed to predict and characterize many important compound attributes (functions) that are required for a compound to become a commercial product. Toxicity profile, mode of action (MoA) classification and identification, product mixture compatibility, environmental impact, and COGS are only some of the critical aspects in the product’s creation process. Traditionally, for cost reasons, these aspects are addressed much later in the development process.
By predicting these factors early on in the virtual screening, and only choosing compounds with the best probability to become products, we accelerate the development process while greatly mitigating downstream risk.
Our discovery method is unique. With a well-defined product concept plan, training data is loaded to a model generator to create a specific activity model. The system then identifies candidates by running the model’s specific predictive algorithms on billions of compounds residing in our massive proprietary database. The output is a shortlist of compound candidates with the predicted desired functionality. The top ranked candidates are then individually analyzed by our scientists and produced to be validated in our plant (in-vivo) laboratories, and, through an iterative process, compounds which effectively fulfil the desired requirements are selected to become leads for future commercial products.
The platform’s simplified custom-built ML/DL models operate on a vast compound database of over 6 billion compounds, which is currently being expanded to 150 billion. This unique data store, together with a novel compound assembler, provides Agrematch unprecedented access to the chemical universe (1060).Furthermore, the ability to screen billions of compounds by our AI4AI predictive system and having high rates of predefined hit discovery, means that we discover and develop compounds in a fraction of the time and costs compared to traditional methods.
Founded in 2017, Agrematch is a data-science product discovery and development company located in Rehovot, Israel, catering to industries that require novel bioactive compounds for their products.
Agrematch’s multidisciplinary team combined computational models, AI, ML, big-data science and extensive biological and chemical know-how to develop agresense™, our revolutionary AI engine for the rational identification of bioactive compounds. To date, the company has proven the value and advantages of its technology by uncovering several active compounds.
Agrematch's AI system and validation platform is especially tailored to identify sustainable and natural compounds that will be safer to the environment. We focus on compounds which can successfully go through the development and registration process to become commercial products for sustainable agriculture.
As a result, we've generated a set of active hits and product leads that comprise the company’s compound pipeline in this domain, with several compound families in various product development phases. Agrematch’s pipeline includes several compounds with weed and disease control properties, which were validated in field trials and are continuing through the development process.
Agrematch's management team is composed of proven entrepreneurs and is backed by a strong advisory board.