Acellera Therapeutics has introduced AceForce 1.0, its groundbreaking neural network potential (NNP) model designed to deliver quantum-level accuracy for predicting atomic interactions, an essential factor in identifying promising drug candidates earlier and more efficiently.

“AceForce 1.0 marks a major leap forward by bringing quantum-like accuracy into everyday drug discovery workflows. Even in this initial release, AceForce 1.0 matches or surpasses state-of-the-art molecular potentials developed over decades. As we expand our training sets, speed up calculations, and refine these AI-driven models, we look forward to empowering scientists to identify promising molecules more quickly, reliably, and affordably,” said Gianni De Fabritiis, Founder and Chief Executive Officer of Acellera Therapeutics.

AceForce 1.0 is a groundbreaking tool in computational chemistry and drug discovery, offering quantum-level accuracy by leveraging a proprietary training set of millions of quantum mechanical (QM) calculations.

This allows it to closely replicate high-level QM methods, providing reliable potential energy surfaces. Its broad applicability spans a wide range of chemical elements and charged molecules, making it a versatile solution for diverse applications in drug discovery and exploration of chemical space.

Efficiency is another key feature of AceForce 1.0, as it can run simulations at nearly twice the speed of previous-generation neural network potentials (NNPs). Its performance has been validated through Acellera Therapeutics' QuantumBind-RBFE platform, where it was benchmarked against publicly available “gold standard” datasets for relative binding free energy (RBFE), further establishing its reliability.

AceForce 1.0 is now available on HuggingFace for non-profit use and demonstration purposes, with tutorials provided for calculations involving small molecules and protein-small molecule complexes. Looking ahead, future iterations of AceForce aim to further enhance accuracy, solidifying its role as a cornerstone in next-generation drug discovery and computational chemistry.