Digital twins will be created thanks to the Gaia AI factory in Krakow

Research on the so-called digital twins, i.e. realistic computer models that reproduce the processes occurring in the human body and the environment, will be carried out thanks to technologies provided by the Gaia AI factory. It is being created at the Cyfronet AGH Academic Computer Center in Krakow.

The following people are involved in the digital twins project: scientists from the Małopolska Center of Biotechnology of the Jagiellonian University and the Sano Center for Computational Medicine.

– The biggest challenge today is not building supercomputers, but designing them so that they can solve real problems. Our task in this project will be to create the so-called digital twins – told PAP Dr. Hab. engineer Paweł Łabaj, prof. Jagiellonian University, from the Małopolska Center of Biotechnology of the Jagiellonian University.

As he explained, a digital twin is an advanced computer model that realistically reproduces the processes occurring in space and time in a specific environment, e.g. in the human intestines. The Krakow project will develop two main types of such models: a digital twin of the intestinal microbiome, which will be worked on by the team of Dr. Tomasz Koscielek from Sano, and the digital twin of the environmental microbiome, developed by the team of prof. Łabaja.

– Such digital twins, based on the infrastructure of the AI ​​factory, will allow us to predict how the body responds to various interventions, including: pharmacological treatment or lifestyle changes – noted the scientist.

He believes that such technology could completely change the way therapies and drugs are developed. It will enable modeling and testing of their effectiveness without the need to conduct some stages of clinical trials in a traditional form, which are currently long-term.

– This may be a breakthrough because it will allow us to create individualized therapies tailored to a specific person – emphasized Prof. Swan.

In the project that will be implemented in Sano, the key factor is the intestinal microbiome. As reminded by prof. Łabaj, more and more research shows that the microorganisms inhabiting our intestines may determine the effectiveness of therapy. An example is the tests of a new anticancer drug conducted several years ago.

– In studies on cell lines and animals, it worked excellently, but in some patients it was completely ineffective. It turned out that the people for whom it didn't work had a specific bacteria in their intestines. It produced a metabolite that completely inactivated the drug. After using the antibiotic and removing the bacteria, the drug began to work properly, said the scientist.

Thanks to digital twins, such relationships will be able to be analyzed already at the therapy design stage. This technology can therefore accelerate the development of the pharmaceutical industry, shorten the clinical trial process and increase patient safety.

According to prof. Łabaja, in the future, it will also be possible to create population models based on it and then personalize them based on data collected from a specific person.

– After appropriate characterization of a given person's microbiome, it will be possible to determine very precisely what products or supplements will be effective for that person. Moreover, in the case of significant dysbiosis (disturbance of the microbiome – PAP), in which the currently administered mixtures of missing bacteria are not always effective, it will be possible to define “pre- and pro-biotic therapy paths”, setting a detailed plan of which preparations to take at what time – emphasized the researcher.

Therefore, in his opinion, the emerging AI factory is an excellent tool for this type of research. – Only now, thanks to the development of computing infrastructure, is it possible to realistically reproduce such complex systems as the microbiome. Previously, we lacked both computing power and high-quality data. Now these conditions are starting to be met, he said.

The data necessary to train the models will come from both scientific projects and the activities of private companies providing microbiome characterization services and collecting this type of information.

The creation of Gaia AI Factory was announced on October 10 by the Ministry of Science and Higher Education and the Ministry of Digitization. As reported, scientists, entrepreneurs and public administration will be able to use the infrastructure.

Katarzyna Czechowicz (PAP)

kap/ agt/ mow/