Researchers develop AI model to predict protein-DNA binding accuracy
The novel technology, developed at the University of Southern California, US is named Deep Predictor of Binding Specificity (DeepPBS).

DeepPBS is a geometric deep learning model specifically designed to predict protein-DNA binding specificity from the structures of protein-DNA complexes.

By allowing scientists to input the data structure of a protein-DNA complex into an online computational tool, DeepPBS provides a faster and more efficient method for determining how proteins interact with DNA.

“DeepPBS is an AI tool that eliminates the need for high-throughput sequencing or structural biology experiments to determine protein–DNA binding specificity,” said Remo Rohs, Professor at the USC’s Dornsife College of Letters, Arts, and Sciences.

This AI tool captures the chemical properties and geometric contexts of protein-DNA interactions to predict binding specificity.

Unlike existing methods that are often limited to one protein family, DeepPBS can predict binding specificity across various protein families.

“It is crucial for researchers to have a universal method that works for all proteins and is not restricted to a well-studied protein family. This approach also enables us to design new proteins,” Rohs added.