- A
- A
- A
- ABC
- ABC
- ABC
- А
- А
- А
- А
- А
Russian Researchers Develop Software to Predict the Location of Elements of the Human Genome
© iStock
Researchers at the HSE Artificial Intelligence Centre have created software for predicting the location of elements of the human genome. The scientists used deep learning methods based on complex data on various human molecular components. The research followed the objectives of the ‘Artificial Intelligence’ federal project of the ‘Digital Economy’ national project.
The research will help genetic laboratories to interpret the results of personal diagnostics more extensively, and will be useful to pharmaceutical companies when searching for targets in medicine development.
Head of the project ‘Artificial Intelligence in Bioinformatics’, HSE University AI Centre
The developed software is a unique solution with a wide range of functionality. It was created using the rapidly developing field of deep learning architectures. Neural network models are easily replaceable with future SOTA models. The aggregation and preprocessing of large volumes of OMIX data is an advantage that will save users weeks, if not months of hard work.
Users can upload OMIX data (a collection of information about various molecular components of the body, such as genes, proteins, metabolites and others) to the server from a web browser. This concept is based on the OMIX prefix, which denotes the study of the global level in the body. Next, the user selects a number of processing parameters: the data type, annotations of genomic functional elements for model training, the complete genome of the analysed type and the architecture of the neural network. After that, the program creates a model and starts the learning process.
At the output, the user receives data on the probability of finding an element in the selected position, statistical analysis of genomic features, and annotation of the studied genome.
Six AI research centres have been created to implement the Artificial Intelligence federal project of the Digital Economy national project in Russia. They operate at Skoltech, ITMO University, MIPT, HSE, Innopolis University and the Institute of System Programming of the Russian Academy of Sciences.