Animol building platform to shake up small molecule innovation
April 7, 2020
Early-stage business Animol is aiming to breathe new life into small molecule drug discovery.
Founded in 2019 by venture capital investor Anterra Capital, Animol is focused on training machine learning algorithms to expedite and improve the screening and discovery process of small molecules for veterinary drugs.
Animol is still in the incubation stage and is currently building its team. It is partnered with artificial intelligence drug discovery organization ZebiAI and has exclusive access to the company’s algorithms for animal health. Animol additionally has access to the drug screening capabilities of biotechnology firm X-Chem.
Phil Austin, co-founder of the start-up and founding partner at Anterra, said: “Animol is being built to address the need for innovation in small molecules for the animal health industry.
“Why small molecules? They’re still the bedrock of this industry. They’re still a proven class of therapeutics that you can reliably discover and develop. They’re inexpensive and robust. “The market for novelty still remains. Whether its isoxazolines, which have driven growth in a number of companies, or whether it’s the poster child for animal health today and growth in the industry (Apoquel), these are still small molecules and there is still space for innovation here.
“But innovation is lagging. There’s very limited investment in foundational innovation – that’s part of the history of the industry. Many of the ‘low hanging fruit’ molecules are now gone. On top of that, as pharma progresses more often into [human] diseases, it’s spending its time and R&D on things that don’t matter so much to the animal health domain.”
Speaking at the recent Animal Health Investment Europe forum in London, Mr Austin said with the separation of veterinary businesses from their parent human pharma firms – the likes of Elanco and Eli Lilly, Zoetis and Pfizer – the connection that would have delivered animal health access to small molecules in the wider organizations is disappearing.
“However, that creates a moment where these [animal health] companies are freer to innovate and the driver for them to innovate in small molecules changes as well because they’re not limited by that relationship,” Mr Austin remarked.
“The heart of the problem is the way we discover small molecules is antiquated. You start with compounds and pure isolates, then you test them one at a time against a target. Although we’ve displaced people with robots, the fundamental approach is the same.”
Mr Austin explained while this approach is fine for human pharma where companies can spend hundreds of millions of dollars chasing billion-dollar drugs, the same cannot be said for animal health. “When you’re chasing $50 million drugs, the cost of this doesn’t really work,” he remarked.
“We’ve paralyzed discovery engines. Diversity is actually highly limited. For example, Bayer Crop Science’s library is about 5.5 million compounds, so when you think about insecticide and parasiticide development, out of 5.5 million compounds available to you, you can only test about 150,000 of those in a discovery campaign. In the world of the chemical universe, that is not even an atom in the ocean. It’s next to nothing.
“The cost of accessing this kind of capability, if you wanted to build a library from scratch, you could easily spend billions of dollars to get to that number of chemical molecules. In areas like crop science you could be looking at spending $100bn just to get the chemical work done, let alone the cost of development.
Phil Austin: “The heart of the problem is the way we discover small molecules is antiquated. Although we’ve displaced people with robots, the fundamental approach is the same.”
“With Animol, it turns out there has been some innovation behind the scenes in pharma in changing the paradigm of how small molecule drug discovery is initiated. The first of these is in approachable DNA-encoded libraries and this really marks the first step-change in pharma discovery, which has now matured. What you can do is now move to testing hundreds of billions of compounds simultaneously.
“On top of that, most recently there’s been a demonstration that the output from a DNA-encoded library screen creates enough information to run machine learning over it. The problem with machine learning is not computer power or algorithms, it’s training data sets. A DNA-encoded library is pretty much the perfect data training set, both in volume and quality of information to train machine learning algorithms to run drug discovery.
“What you end up with when you move to algorithmic development is a very powerful platform. Whether you’re looking at known chemical classes and trying to find new chemical spaces, new targets or known targets. Or you can get into smarter re-purposing, so rather than using clinical data as the benchmark to repurpose a human pharma drug, you could actually go back one step further and constrain your search in small molecules to compounds that are known [to be safe] but do not necessarily have the clinical data.
“So, algorithmic development enables you to explore not just broader space but actually the constrained spaces more dynamically, so you can focus on products that have greater development capability built into them.”
Mr Austin believes Animol’s access to DNA-encoded library capability through its partners is the best in the industry. He claims the start-up has already started having conversations with animal health stakeholders about their drug discovery operations and said although the firm is wholly funded by Anterra, it is also discussing funding with other biotech investors.