AI reveals the role of junk DNA in Alzheimer's disease

AI has helped scientists prove that junk DNA plays a key role in Alzheimer's disease. Artificial intelligence and CRISPR have made it possible to identify hidden DNA switches in brain cells, which is changing the approach to the study of neurodegenerative diseases.

Artificial intelligence is becoming not just a tool for data analysis, but a full-fledged participant in scientific discoveries, writes xrust. A recent study by an international team of scientists showed that it was AI that helped decipher the functions of so-called “junk” DNA — a large part of the genome that was considered useless for decades. It turned out that these regions are directly related to the development of Alzheimer's disease and could become the basis for new diagnostic and treatment methods.

AI is changing the understanding of “junk” DNA and Alzheimer's disease

For a long time, it was believed that genes encoding proteins play a major role in the functioning of the body. However, they occupy only about 2% of the human genome. The remaining 98% — non-coding regions — have traditionally been called “junk” DNA.

AI models trained on large sets of genetic data have provided new insights into these areas. Machine learning algorithms have revealed that non-coding DNA contains thousands of regulatory elements that control how genes turn on and off in brain cells.

AI and the search for hidden DNA switches in astrocytes

Astrocytes are auxiliary cells of the brain that support the functioning of neurons. Increasing evidence indicates their involvement in the development of Alzheimer's disease.

Key facts of the study

• The study was carried out by scientists at the University of New South Wales (UNSW, Australia)
• The work was published in the journal Nature Neuroscience on December 18
• About 1000 potential DNA switches were analyzed
• The activity of approximately 150 enhancers

Enhancers are sections of DNA that can be located at a great distance from genes, but still control their activity. Without the help of AI and automated analysis, identifying their impact would be almost impossible.

AI, CRISPRi and single-cell sequencing

Scientists combined several advanced technologies for a large-scale experiment.

Methods used

• CRISPRi — “soft” disabling of DNA sections without cutting the genome
• Single cell RNA sequencing
• AI algorithms for processing and collating data

According to the study's lead author, Dr. Nicole Green, it was automated analysis that made it possible to link the shutdown of enhancers to changes in gene activity.

“AI has helped us see patterns that cannot be detected through manual analysis. Most of the functional enhancers turned out to be associated with genes involved in Alzheimer's disease,” notes Green.

AI as a tool for interpreting genetic risks

Project leader Professor Irina Voineagu emphasizes: the results are important not only for the study of Alzheimer.

“When we look for the genetic causes of complex diseases — from diabetes to mental disorders — mutations are more often found between genes, rather than within them. AI allows us to understand which of these regions actually have functional significance.»

The study created a catalog of validated regulatory elements that can be used to interpret data from GWAS studies (genome-wide association studies).

AI and training of new DNA prediction models

The resulting dataset has become a valuable resource for AI developers. According to Professor Voineagu, it is already being used to test deep learning models.

Who is using the research data

• DeepMind team (Google)
• AlphaGenome model developers
• Computational biology specialists

AI models learn to predict which regions of non-coding DNA are active switches without expensive laboratory experiments. This could shorten years of research.

AI and the promise of gene therapy

Because many enhancers work strictly in certain cell types, they become a promising target for targeted gene therapy.

Professor Voineagu reminds us that there is already a precedent: the first approved drug for the treatment of sickle cell anemia targets a cell-specific enhancer.

AI plays a key role in this context — it helps to accurately determine which regulatory elements can be safely used in therapy.

Background: why this discovery important

Alzheimer's disease remains one of the most difficult to understand and treat. Until now, most research has focused on individual genes and proteins. A new approach based on AI and analysis of non-coding DNA is changing the very paradigm of neurobiology.

Exclusive commentary from an independent expert, molecular biologist Sergei Kovalenko:

“This study is an example of how AI ceases to be an auxiliary tool and becomes a co-author of scientific discoveries. Without machine learning algorithms, such a volume of data would simply be impossible to interpret.”

Xrust AI reveals the role of “junk” DNA in Alzheimer’s disease