Researchers from HSE University and Sber Artificial Intelligence Lab are using AI to predict the location of DNA fragments in the genome which can flip over and form a mirror structure known as Z-DNA. The scientists have found that these DNA fragments overlap with known mutations which can cause severe hereditary diseases and impact a person's health, height, weight, cholesterol levels, and even determine their hair colour. The study findings have been published in Life Science Alliance.
Z-DNA represents the mirror form of a DNA molecule. While the classical and most common structure of DNA is a double helix twisted to the right, Z-DNA is a helix twisted to the left. Special sections of DNA known as 'flipons' have the ability to alter the typical shape of a DNA molecule into its mirror image. In addition to flipping the molecule, they can also modify the readout of genomic information.
Researchers from the HSE International Laboratory of Bioinformatics, together with colleagues from the Sber Artificial Intelligence Lab, analysed the genomic locations of Z-DNA flipons to gain a deeper understanding of the role of inverted DNA structures in the human body.
Using a neural network algorithm pre-trained on experimental data, the scientists achieved a high level of accuracy in predicting the locations of flipons across the genome. These locations were found to coincide with genes whose mutations can give rise to severe hereditary diseases. Inverted DNA has the capacity to alter the genetic program.
Specifically, the researchers have established a connection between Z-DNA and congenital diseases including hemoglobinopathy, osteogenesis imperfecta, Waardenburg syndrome, and others. Furthermore, when a mutation coincides with the regions of Z-flipon activity, it can influence an individual's traits and the processes occurring in their body.
The experimental data reveals those specific DNA regions which exhibited an alternative conformation at the time of the experiment. However, there are numerous other sites that have the potential to adopt alternative forms – more than can be observed in any particular experiment. The developed algorithm maps all sites capable of altering their conformation and enables biologists to explore the relationship between flipons and other genetic mechanisms.
This work is an important step towards a better understanding of the role of alternative DNA structures in the human body. We were able to establish a resource of Z-flipons, which can be used for testing various hypotheses in the future. The flipon map will aid in identifying the relationship between mutations and various clinical conditions, leading to the discovery of new approaches to treating serious diseases.