News

Artificial intelligence as key to predicting the risk of type 2 diabetes

- Projects, Research

Credit: DGLimages - Getty Images.

Research teams from the Barcelona Supercomputing Center-Centro Nacional de Supercomputación (BSC-CNS), the University of the Basque Country/Euskal Herriko Unibertsitatea (UPV/EHU), and the Germans Trias i Pujol Research Institute (IGTP) are working on a project that uses artificial intelligence to predict and prevent type 2 diabetes. This initiative is part of the Complementary Plan for Biotechnology Applied to Health, coordinated by the Institute for Bioengineering of Catalonia (IBEC).

In developed countries, a significant portion of healthcare focuses on treating chronic non-communicable diseases like type 2 diabetes, one of the most common. Research has intensified efforts to find treatments and protocols that can predict and prevent these diseases.

Developing predictive models for type 2 diabetes continues to pose significant challenges today. It is necessary to integrate a comprehensive view that considers all elements we are exposed to through our diet, lifestyle, and the environment in which we live and work, as well as internal biological factors such as metabolism, genetics, and epigenetics, known as the exposome.

Numerous studies have been conducted to identify variations in specific genes in the hope that they will be useful as therapeutic targets and for predicting the risk of developing the disease. However, the limited number of people studied from specific populations and the difficulty in integrating other variables beyond genetics, such as epigenetic markers, mean that the validity of the predictive models is still limited.

The project "AI4T2D: Development and Implementation of Integrated Artificial Intelligence Models for the Prediction of Type 2 Diabetes Risk" uses artificial intelligence to build self-learning based models that identify epigenetic, clinical, and environmental determinants. In this way, researchers hope to predict the risk of the disease and classify patients by prognosis and response to existing treatments.

Rafael de Cid, leader of IGTP's strategic project GCAT|Genomes for life and co-PI of the project, explains: "The combined models we aim to develop will not only propose new therapeutic targets but also identify predictors that allow the implementation of personalised clinical protocols for each individual, as well as environmental health at a population level". The current project will analyse a clinical sub-cohort of 400 GCAT volunteers. By comparing methylation profiles analysing about 900,000 CpG sites across the entire genome, from healthy volunteers and those diagnosed with type 2 diabetes, to identify the epigenetic mechanisms involved in the onset and progression of the disease.


The project started last June and will last for 22 months. It falls under Action Line 2 of the Plan Complementario de Biotecnología Aplicada a la Salud, which focuses on the acquisition and study of extensive experimental databases to enable the physiopathological characterization of the population. The project has received co-funding from the Spanish Ministry of Science and Innovation with funds from the European Union's NextGenerationEU, the Plan de Recuperación, Transformación y Resiliencia (PRTR-C17.I1), and the Generalitat de Catalunya.

The entities involved in this project are the Barcelona Supercomputing Center-Centro Nacional de Supercomputación (BSC-CNS), the University of the Basque Country/Euskal Herriko Unibertsitatea (UPV/EHU), and the Germans Trias i Pujol Research Institute (IGTP).