Machine learning and structural biology company Atomic AI announced that it has created the first large language learning model that leverages chemical mapping data to optimize RNA therapeutic development by predicting the structure and function of RNA. The company details the development of its proprietary technology, called ATOM-1, in a preprint published on bioRxiv.
The development of mRNA-based COVID-19 vaccines has underscored the potential of RNA-based and RNA-targeting therapies in addressing a broad swath of health conditions, including infectious diseases, cancer, and neurodegenerative disorders. But existing challenges to designing and discovering RNA therapeutics arise from a lack of data essential for predicting the structure and function of RNA. Current methods, such as utilizing animal models for in vivo studies or employing cryo‐electron microscopy (cryo‐EM) for determining 3D RNA structure, are both difficult to master and time-consuming. The scarcity of high-quality RNA data has hindered the optimization of important RNA therapeutic attributes, including stability, toxicity, and translational efficiency, due to the absence of reliable “ground-truth” data.
“ATOM-1 enables the prediction of structural and functional aspects of RNA as well as key characteristics of RNA modalities, including small molecules, mRNA vaccines, siRNAs, and circular RNA, to aid in the efficient design of therapies,” said Manjunath “Manju” Ramarao, PhD, chief science officer of Atomic AI. “Our goal is to create a streamlined drug discovery process to advance our own pipeline and work with partners to help validate their RNA targets and tools, to ultimately get needed therapeutics to patients quickly and more efficiently.”
As detailed in the preprint, the Atomic AI team created their ATOM-1 platform by building an in-house dataset of using wet-lab assays, that included data from millions of RNA sequences and more than a billion nucleotide-level measurements. These data were then used to train ATOM-1 which provides more detailed insight of RNA which can then be used to optimize the properties of RNA modalities for therapeutic development.
“By building large datasets based on RNA nucleotide modifications and next-generation sequencing, the team at Atomic AI has created a first-of-its-kind RNA foundation model,” said Stephan Eismann, PhD, founding scientist and machine learning lead at Atomic AI. “We are excited about how broadly applicable our model is to other aspects of RNA research, and its potential for optimizing various properties of RNA-based medicines, such as the stability and translation efficiency of mRNA vaccines or the activity and toxicity of siRNAs.”
According to the company, its model stands apart from others in previously published studies based on its ability to more accurately predict secondary and tertiary RNA structure. The Atomic AI teams retrospective analysis of other tools for vaccine design, for instance, it showed it outperformed existing models that predict in-solution mRNA stability. This allows for the prediction of different properties of the RNA, as well as other features of RNA-based therapies with a limited amount of data.
“Over the last two and a half years, we’ve been purposefully designing and collecting data to train our foundation model,” said Raphaël Townshend, PhD, founder and CEO of Atomic AI. “Through machine learning and generative AI, we now have a unique opportunity with ATOM-1 to predict RNA structure and function with high precision by tuning it with just a small amount of initial data points.”
