---
title: "AlphaFold Makes Gene Therapy Safer: AI Reduces Error Risk from 28% to 5%"
description: "AI AlphaFold helped make gene therapy safer: scientists reduced the risk of errors in DNA editing from 28% to 5%. 🧬🤖 This is a breakthrough for personalized medicine."
date: 2026-07-28T16:16:00.000Z
lang: en
url: https://xab.info/en/posts/alphafold-makes-gene-therapy-safer-ai-reduces-error-risk
tags: [alphafold, gene-therapy, nature, ai]
publisher: "XAB.info"
---

# AlphaFold Makes Gene Therapy Safer: AI Reduces Error Risk from 28% to 5%

![Abstract visualization of a DNA helix with colorful light lines, symbolizing AlphaFold AI's role in gene therapy](https://xab.info/media/2026/07/28/alphafold-sdelal-gennuyu-terapiyu-bezopasnee-ii-snizil-risk-oshibok/alphafold-sdelal-gennuyu-terapiyu-bezopasnee-ii-snizil-risk-oshibok-1.webp)

Gene therapy, capable of changing the course of treatment for hereditary diseases, has received a powerful boost. Scientists have been able to significantly increase the safety of procedures by leveraging the capabilities of artificial intelligence. Thanks to the AlphaFold system, researchers have learned to identify and eliminate potential errors that were previously extremely difficult to predict.

New data has been published in the prestigious journal Nature. As reported by RBK-Ukraine, an international team of scientists managed to solve one of the main problems of modern genetic engineering — the risk of off-target effects on DNA.

### How molecular "scissors" work

The essence of modern gene therapy can be compared to the work of a jeweler or a surgeon at the molecular level. Special proteins act like "scissors": they scan the human genome, find the damaged section, and cut it out, replacing it with a corrected version.

However, the human genome is colossal in scale. In this vast space, "molecular scissors" sometimes go off course. Errors are rare, but they are inevitable. If a protein cuts the wrong gene, it can lead to dangerous mutations and serious side effects for the patient.

### The role of artificial intelligence

To eliminate this risk, scientists turned to the help of AlphaFold — a system capable of modeling complex molecules with atomic precision. Artificial intelligence allowed them to look inside the process and understand the mechanics of errors.

Analysis of thousands of variants showed that the problem lies in the excessive flexibility of the editor protein. It can accidentally latch onto DNA sections that resemble the target but are "foreign." AI accurately determined which specific details of the protein structure are responsible for these false bindings.

### Results: from 28% to 5%

With hints from the algorithms, researchers modified the protein's structure. The experimental results exceeded expectations:

    - The updated system retained the ability to accurately "repair" the necessary genes.

    - The number of dangerous errors on adjacent DNA sections dropped from 28% to 5%.

This discovery is of fundamental importance for medicine. The new method allows for flexible tuning of gene tools for a specific patient and the specifics of their treatment. Experts note that this brings us closer to the era of personalized medicine, where complex genetic diseases can be treated without threat to life.

"This discovery will allow the creation of safe and precise medicine for each specific patient, minimizing the risk of complications," the authors of the study summarized.