Imagine controlling genes with the precision of a light switch— controlling specific parts of the body as you will. That’s what the researchers at Yale, Jackson Laboratory, and the Broad Institute aimed to achieved with a new AI tool called CODA (Computational Optimization of DNA Activity). This newly designed AI tool creates custom DNA “switches” that control when and where genes turn on or off, which could greatly improve gene therapy.
What is CODA?
CODA is a groundbreaking AI that can design unique DNA sequences called “regulatory elements” to control genes with pinpoint accuracy. The designs generated by this AI tool can act as templates to make sequences which act like tiny on/off switches, targeting only specific types of cells. For example, imagine being able to turn on genes in brain cells affected by Parkinson’s disease without affecting other parts of the body.
How CODA Works
CODA uses advanced machine learning to study how natural DNA elements control genes. By analyzing large amounts of data from human blood, liver, and brain cells, it creates new DNA switches that work in specific cell types. When tested in lab settings and animals, CODA’s synthetic DNA sequences were highly accurate, turning on specific genes in targeted cell types—like neurons in the brain.
Why CODA Matters for Gene Therapy
One of gene therapy’s biggest challenges has been finding ways to target only the diseased cells while avoiding healthy ones. Early gene therapies sometimes affected healthy cells, leading to unwanted side effects. CODA’s ability to create precise DNA switches could make gene therapy to safer and more effective.
Future Goals for CODA
Researchers aim to further develop CODA to create even more precise switches for a variety of cell types. This progress could open the door to new, highly targeted gene therapies that bring effective treatment to patients with fewer risks.
Created: Oct 24th, 2024
Citations:
Tompa, R. (2024, October 23). Generative AI designs DNA sequences to switch genes on and off. Yale School of Medicine.
Comments