“We are essentially just going in the direction that has been so successful for the pharmaceutical industry,” AgPlenus CEO Dan Gelvan says of his company’s approach to develop crop protection products.
This means taking a more targeted approach to product development, rather than creating, say, one herbicide that kills 30 weeds and a lot of beneficial plants in its wake.
AgPlenus, a subsidiary of computational biology company Evogene, employs an AI-based computational platform to discover novel modes of action that can target specific pests or diseases.
The company recently announced a new mode of action active against Zymoseptoria tritici, the fungal pathogen responsible for Septoria tritici blotch that impacts wheat crops, particularly in Europe. In recent months, AgPlenus has also struck partnerships with major agricultural players including Corteva and Bayer.
“At the end of the day, what sets us apart is being a computational company,” says Gelvan. “We come in with a very, very interesting set of of tools that help us to be innovative and to generate interesting products that can actually become pesticides.”
AgFunderNews (AFN): Tell us more about the “target-based” approach to developing pesticides.
Dan Gelvan (DG): This is very different from what the [crop protection] industry does today, where people are looking for broad scope: “Let’s see if we can find something that kills 30 weeds at the same time, etc, etc.”
When you do target-based, you essentially drive yourself into something that looks much more like the development of a drug.
We start by building a target product profile—a TPP. That essentially means we have to decide on a pest that we want to kill in a certain set. This is very specific.
[Then] we build a list of diseases that we think can [lead to] “blockbuster” pesticides. [The product] has to be big enough so we can get a share that is big enough to cover our expenses, and the partner needs to have a big enough share so what is left for them makes it economically viable.
So we’re looking for major problems, and one of the things we honed in on was on Septoria, mainly because it’s a big problem in Europe.
It’s really significant that this is part of what we put into our target product profile—how big a problem is for the farmer. At the end of the day, your value proposition is not going to be about being a nice guy. Your value proposition is [whether you] solve a real problem for the farm.
Septoria is a real problem, because at least in winter wheat, farmers can lose 50% to 90% of the crop.
AFN: What convinced you to take this target-based approach versus something broader?
DG: We [the biologicals industry] are lagging behind the pharmaceutical industry, which also used to do very broad things. It used to talk about “anti-cancers.” Today, we talk about cancers at the level of a receptor or a single mutation, and that’s not a broad scope. No one talks about breast cancer in the broad sense anymore; it’s always [about] what kind of receptors are active, down, etc., what mutations [there are].
We are essentially just going in the direction that has been so successful for the pharmaceutical industry, and that is to be more focused on the organism you want to kill, or the pest you want to kill. Try to learn as much as you can about it and try to predict the optimal ground to kill this pest.
The farmer needs more solutions in order to overcome [pest and disease] resistance, because often when we start to see resistance, the application [of the crop protection solution] rate goes up. Maybe they do it more frequently. Maybe they use a higher concentration. But what they really need is to start to do integrated pest management. But with IPM, you need many different targets and different modes of action. So that’s where we come in. Our strategy is that everything we do will come in with a new target or a new mode of action.
AFN: Is there a possible broader application here too?
DG: That we look for a fungicide that works on a certain fungus in a certain setting doesn’t mean it doesn’t have a broader application.
At the very least, it’s going to kill the whole family of Septoria, not only the tritici. We also have already started to see that it kills other fungus.
But the work that we do is focused on one pest, and this is what sets us very much apart from the industry standards.
AFN: Why are partnerships so critical to what you do?
DG: It’s crucial for us to find partners, because at the end of the day, what sets us apart is being a computational company. When we collaborate with bigger companies, we take the predictive part of it.
A typical collaboration agreement would be, we start out with collaborating for a number of years. We have optimization tools, and we will computationally try to suggest more if potency is the problem, if cost of goods is the problem, if sustainability targets are off, etc.
Our latest tool is really mind boggling. It’s sort of a chat GPT engine. It was developed by Evogene and it is an LLM that has learned what a molecule looks like, what herbicides look like. For example, we take all the results of what we did in the whole process and feed them into the system, and it starts to suggest molecules.
Why is it mind boggling? Because we’ve been implementing it now in a couple of projects, and we have probably in excess of 90% novelty now. Novelty means the model suggested a molecule that it has never seen.
We have collaborations because they generate new chemistry. New chemistry is new IP, new patents and better protection, and downstream that’s going to make better money.
Source link
Author Jennifer Marston
