Study seeks to identify existing remedies for different diseases

A survey evaluated millions of scientific articles to look for possible as-yet-unknown links between diseases of different origins and types.

The findings are expected to contribute to innovations in drug repositioning — when a drug is used in a treatment different from the one for which it was initially developed.

The research was published in the scientific journal iScience and is signed by researchers from different Brazilian institutions, such as UFMG (Federal University of Minas Gerais) and USP (University of São Paulo).

Helder Nakaya, one of the authors of the study and a senior researcher at Einstein, explains that one of the reasons for starting the project was the fact that the amount of published articles is enormous, which prevents a scientist from being able to access all the knowledge available to a illness.

“This is very frustrating because what we want is to get all the knowledge about that disease and be the main specialist for it”, he says.

In total, 30 million articles published between 1990 and 2018 were analyzed, covering 99 diseases and more than 3,700 genes. To access all this content, artificial intelligence was used to identify relationships between genetic markers, diseases and the drugs that were used in their treatments.

“What the [método] did was automatically read millions and millions of articles, create a network of relationships between genes, diseases and drugs, so you actually had a network of knowledge connections”, he says.

Nakaya explains that, in the study, some parameters were considered to avoid errors in the analysis of the articles. For example, a relationship between genes, diseases or drugs was only considered when it was found in two independent publications, not to consider studies that were still very incipient.

From this network arising from the analysis of publications, it is possible to observe the genes that are related to a disease and the drugs that treat this disease. It can also be compared with other diseases that have similar genetic markers and eventually test whether the drug used to treat one disease can be repositioned for another.

Even with the aim of generating this map of relationships, the research does not seek to establish the repositioning of drugs. “This is an initial work that serves to point out ‘look people, try to use this drug to treat this disease'”, he explains.

According to Nakaya, it is important that further research be done to confirm or not the safety and feasibility of reusing a drug for another disease.

In this way, the researcher mentions that there are already some initiatives that observe the mapped network to propose other investigations. One of these cases is a drug that, if repositioned, can act against schizophrenia and which is being studied by scientists at Unicamp (University of Campinas).

“The idea is to test a drug used against arthritis to validate its potential use in the treatment of schizophrenia”, says Daniel Martins-de-Souza, professor of biochemistry at Unicamp.

He explains that the starting point for this research was the associations of genetic markers between the two diseases — the same thing observed in Nakaya’s research. The hypothesis is that the drug has an anti-inflammatory function with schizophrenia and also acts on atrocytes, which are cells associated with the disease.

At Unicamp, researchers are carrying out in vitro studies on this cell type and, according to Martins-de-Souza, the initial result “has been very promising”.

Other analyzes of the drug that have been carried out in animals at the Faculty of Medicine of Ribeirão Preto, USP, have had preliminary outcomes similar to the in vitro model.

Although the initial analyzes are positive, several more steps are still needed to prove the effectiveness and safety of the drug for another disease.

If the possibility of use for schizophrenia is proved, the biochemistry professor at Unicamp explains that patients can have two benefits: fewer side effects and greater effectiveness in some cases.

This concern to arrive at a treatment that causes fewer problems for patients and is more effective is also something that Nakaya addresses in her research.

“A network [de conexões de genes, doenças e fármacos] it’s important because you can see which are the specific regions that can be affected by drugs without destroying and without affecting the whole network”, he says.

At the moment, he is already involved in other research of a similar nature, but now focused only on viral diseases. “It’s an analysis similar to this one, only for all human viruses, including obviously Sars-CoV-2,” he says.

The expectation is similar: to identify the genes associated with the diseases caused by these pathogens, which may result in new means of treatment.

“If you know which genes are affected by the different viruses, if [eles] are respiratory or if they are other types of viruses, what are the mechanisms associated with them, so we can understand which genes are generic, that is, that any viral infection increases these genes”, he explains.