AI-driven computation will enhance speed and accuracy of drug discovery.
Open-source machine learning framework Nvidia BioNeMo has been added to the Sapio Sciences Informatics Platform, a configurable LIMS for laboratory automation.
The integration of the BioNeMo platform aims to accelerate drug discovery by leveraging AI frameworks, pre-trained models and generative AI tools to streamline the identification of potential drug candidates and improve target selection accuracy.
“AI innovation is advancing rapidly, but scientists are often forced to navigate fragmented tools with complex interfaces, slowing down research,” said Sapio Sciences founder, CEO and CTO Kevin Cramer.
“Our integration of Nvidia’s powerful AI-driven tools directly into the Sapio platform enables researchers to apply AI seamlessly into their experiments.”
The BioNeMo platform provides scientists with a framework for training and deploying large biomolecular language models at supercomputing scale.
The addition of the BioNeMo platform helps to bring AI-driven computational drug discovery directly into Sapio Electronic Lab Notebook (ELN).
With Sapio ELN, researchers can access BioNeMo NIM microservices to identify and optimise drug candidates with AI-driven molecular modelling with a range of NIMs. This allows researchers access to AI-driven tools without extensive setup, streamlining workflows.
“Integrating BioNeMo into Sapio’s AI-driven research platform gives scientists access to advanced generative AI models for drug discovery,” said Nvidia director, digital biology Anthony Costa. “With AlphaFold2, MoIMIM, and DiffDock NIMs, researchers can predict, optimize, and validate drug candidates with greater speed and accuracy.
AlphaFold2 NIM helps to predict accurate 3D protein structures, while MoIMIM NIMenables the design and optimisation of small molecules. DiffDock NIM is an AI-powered docking model developed by MIT.
AI-driven molecular simulations are available within a single, unified workflow, allowing researchers to optimise their processes and easily transition from discovery to development. These silico studies enable scientists to, for example, generate novel candidate molecules early in the research process and test their docking with a target protein.
“Through this work, we are removing inefficiencies and equipping scientists with the tools to rapidly generate, analyse, and visualise both chemical and biological results,” added Cramer.
“This collaboration is a major step toward making AI an integral part of the drug discovery process, helping researchers make faster, data-driven decisions and improving research outcomes.”
