DeepMind, a Google company focused on developing Artificial Intelligence (AI), uses new tools that make it possible identify changes in human DNA, and researchers believe they have identified 89% of the major genetic mutations. In this way, diagnosis and the search for better treatments can be accelerated.

Using AI to decipher DNA

DNA consists of four blocks of chemical compounds called adenine (A), cytosine (C), guanine (G) and thymine (T). In humans, when the embryo develops, the sequence of these letters is read to produce proteins, which are the building blocks of the cells and tissues that make up various parts of the body.

If the order of these letters is incorrect, for example due to a congenital abnormality, cells and tissues in the body do not form properlywhich can cause disease.

Last year, Google DeepMind AI solved the structure of almost all proteins in the human organisman important advance in a new tool called AlphaMissense for determining whether DNA letters will produce the correct structure or, if not, labeling them as potentially disease-causing.

Pushmeet Kohli, a scientist at Google DeepMind, commented this way Their new model successfully classified 89% of genetic mutationseither benign or disease-causing.

Instead of searching for potentially disease-causing regions through the billions of blocks of chemical components that make up DNA, Kohli explains that “researchers can now focus their efforts on new regions they didn’t know about” and that, with their new discovery, is that they are not aware of the potential disease-causing regions. the device, has been identified as “potentially causing disease”.

The new DeepMind system has been tested by Genomics England and described in an article in Science magazine.

Ellen Thomas, deputy medical director of Genomics England at the UK’s National Health Service (NHS), “this new tool really brings another perspective to the data.” In turn, he said: “This will help clinical scientists understand genetic data so that it is useful to patients and their clinical teams.”