Researchers reveal how diseases interact at the genetic level

The human body is a complex and interconnected system, where alterations caused by one disease can promote the onset of others. This tendency for certain diseases to occur together, beyond what would be expected by chance, is called co-occurrence. Thus, although there are diseases with widely known co-occurrence in certain groups of patients, such as Crohn's disease and the development of ulcers, many of the molecular mechanisms that would explain them were, until now, unknown.

A study by the Barcelona Supercomputing Center – Centro Nacional de Supercomputación (BSC-CNS) analysed molecular data from more than 4,000 patients and 45 diseases using a newly developed computational method. This research represents the largest effort to date to scientifically explain the clinical associations between diseases. The results show that 64% of medically known connections are related by similarities in gene expression, and provide relevant clues about the biological mechanisms that link them.

Using RNA sequencing data, a technology that allows researchers to read which genes are active in each patient, they were able to trace the relationships between complex diseases, observing positive interactions in which the presence of one disease favours the onset of others, as is the case with asthma and Parkinson's disease; or negative interactions, in which some patients with one disease appear to be protected against the development of others, such as between cancer and neurodegenerative diseases like Huntington's.

We have known for years that patients with Huntington's disease develop fewer solid tumours, such as lung or breast cancer, than would be expected by chance. This study provides a possible molecular explanation for this phenomenon, revealing that many of the biological processes associated with Huntington's disease follow pathways opposite to those of cancer. We can now investigate these mechanisms and learn from them."

Beatriz Urda, researcher at the BSC and lead author of the study

The results indicate that the immune system acts as the central axis of these interactions, as common alterations in immune pathways have been detected in 95% of clinically related diseases. Furthermore, the study identifies new possible associations, such as Down syndrome and lupus, which could improve the diagnosis of certain diseases and the development of new therapeutic strategies.

Patient groups and personalised medicine

However, many of these co-occurrences have only been detected by dividing individuals with the same disease into subgroups according to their molecular profiles. i.e., by grouping patients who have the same genes active or inactive. For instance, certain subgroups of breast cancer patients have been observed to exhibit molecular connections with autism or bipolar disorder, while others demonstrate a negative interaction that could potentially protect them from multiple sclerosis.

"The study has revealed that many associations only emerge in certain patients, which would explain why two people with the same disease can have completely different clinical trajectories. This approach allows us to identify potentially underdiagnosed associations and propose molecular mechanisms to explain clinical links that have been poorly understood until now," Urda pointed out.

This new methodology could also be particularly useful for studying rare diseases, "which are often more difficult to characterize due to the scarcity of clinical data. Despite these limitations, the computational method has a capacity for detecting interactions comparable to that of more common diseases and could open the door to a better understanding of these minority pathologies," explains Alfonso Valencia, ICREA professor, study leader and director of the Life Sciences Department at the BSC.

This research not only helps explain clinical phenomena observed for decades, but also opens new avenues for anticipating which diseases a patient might develop and for adapting treatments in a more preventive and personalized way. It thus underscores the potential of integrating clinical and genomic information to better understand diseases, not as isolated entities, but as part of a system interconnected by their underlying molecular characteristics.

Following the collection of all relevant data and the study of all interactions, the BSC scientific team launched a web resource open to the public and the scientific community. This platform allows interactive exploration of the positive and negative associations between numerous diseases, as well as the possible molecular mechanisms behind each link.