Bringing a new drug to market is a complex process associated with high costs, long timelines, and a low success rate. However, the field of drug discovery is evolving rapidly, transitioning from reliance on the chance to rational design and high-throughput screening. The next frontier in this revolution is artificial intelligence (AI), holding potential to accelerate and streamline the process.
“While existing AI efforts in drug discovery have shown promise, most heavily rely on existing datasets needed for training,” says Vardan Harutyunyan, Denovo Sciences’ co-founder and Chief AI Officer. “With only several thousand FDA-approved drugs, the data is both scarce and challenging/expensive to acquire. Moreover, these approaches are limited by the biases and constraints inherent in the data itself.
Denovo Sciences employs a Deep Reinforcement Learning approach (DRL) to train on physico-chemical simulations, enabling the generation of novel therapeutics without relying on training datasets. “This technology empowers Denovo to identify therapeutics for challenging targets, neglected diseases, and unconventional biological targets,” Vardan says. “The technology has already been validated in vitro and in vivo across multiple diseases.”
In the scope of the ELISE project, Denovo plans to enhance their platform by creating a new molecular representation format that harnesses the structural and 3D information of ligands, proteins, and their interactions. The new representations aim to more accurately model real-life interactions, ultimately improving the robustness and enhancing the quality of generated molecules.
At the moment, Denovo Sciences has several internal drug discovery pipelines and multiple ongoing partnerships with industry and academia. The long-term plan is to expand the internal pipeline and target a broader range of diseases, with the goal of making undruggable diseases a thing of the past one molecule at a time.
Find out more about Denovo Sciences and their work here and follow them on LinkedIn here!
More about the ELISE 2nd Open call and awarded companies is available here.
