SUNY Poly Assistant Professor of Computer Science Dr. Amirhossein Manzourolajdad has published groundbreaking research on RNA inverse design, introducing a novel approach using Relational Graph Neural Networks (RGNNs) to improve computational RNA sequence prediction. The study titled, “Secondary-Structure-Informed RNA Inverse Design via Relational Graph Neural Networks,” was published in MDPI’s Non-Coding RNA and offers significant advancements in the field of RNA-based therapeutics and synthetic biology.

RNA inverse design is a critical area of research with applications in gene therapy, drug development, bio-sensing, and synthetic biology. Dr. Manzourolajdad’s work focuses on improving the design of RNA regulators, such as riboswitches, which are complex molecular structures that control gene expression. His approach enhances the accuracy of RNA sequence prediction by incorporating alternative RNA structures into machine learning models.

Using a relational geometric graph neural network, the research introduces a secondary-structure-informed approach that models RNA molecules as 3D geometric graphs, integrating primary, secondary, and spatial interactions. The results demonstrate a higher native sequence recovery rate than previous methods, significantly improving the ability to predict RNA sequences that fold into desired structures. This breakthrough could lead to more efficient design strategies for RNA-based drugs, molecular tools, and therapeutic interventions.

Dr. Manzourolajdad is pictured above with SUNY Poly graduate students Nitin Manne and Pradeepthi Rangineni, who are exploring this work further and recently presented a poster at a research day at Utica University.