Researchers flip genes on and off with AI-designed DNA switches

AI-Designed DNA Switches for Precise Gene Control: Researchers created synthetic cis-regulatory elements (CREs) using artificial intelligence to control gene expression in specific cell types, enabling targeted gene regulation.

High Specificity in Targeting Cell Types: These new synthetic CREs show remarkable specificity, activating genes only in the desired cell types (e.g., liver, brain, or blood cells) without affecting other tissues, overcoming a major limitation in traditional gene therapies.

Deep Learning Model for Predicting CRE Function: By training a deep learning model on thousands of DNA sequences, researchers identified new DNA patterns that predict CRE activity, effectively "decoding" the grammar of these regulatory elements.

CODA Platform for CRE Design: The team developed CODA (Computational Optimization of DNA Activity), an AI platform capable of designing CREs with customized characteristics, such as activating genes in one organ but not others.

Testing and Validation in Living Models: Synthetic CREs were successfully tested in zebrafish and mice, showing controlled gene expression in specific tissues, demonstrating the potential for precise therapeutic applications and advancing gene therapy possibilities.