The language of virus-host interactions

Helmholtz funds AI project in Würzburg and Munich

The COVID-19 pandemic made it abundantly clear that a better understanding of emerging viruses and their variants is of great importance to society. Mathias Munschauer from the Helmholtz Institute Würzburg (HIRI) and Annalisa Marsico from Helmholtz Munich are tackling this task: with the help of artificial intelligence (AI), they want to decipher how viral ribonucleic acids (RNAs) interact with host proteins. Helmholtz is funding the research project as a “Helmholtz AI Project”.

In 2020—shortly after the outbreak of the pandemic—the lab of HIRI group leader Mathias Munschauer created the first RNA-protein interactome of SARS-CoV-2, the virus that triggers COVID-19. “We now have a list of all the proteins that bind to the viral RNA genome. But we don't know where in the genome they bind and why,” says Munschauer. So which RNA sequence or structure is responsible for the interaction? This is where Annalisa Marsico, group leader at Helmholtz Munich, comes into the picture. She has developed AI-based methods that can predict which proteins will bind to which RNA region. The Munschauer lab generates the necessary experimental interaction data in cells infected with SARS-CoV-2. The Marsico group then uses this data to train the models, predict the binding sites of the proteins in the context of infection and predict potential novel variants of concerns that alter host protein binding.

The research project aims to determine for each protein whether it will bind SARS-CoV-2 RNA and to map exactly where this interaction takes place—even without experimental data. In the future, this method could be applied to other viruses. For example, before a pandemic outbreak, it may be possible to determine which inhibitors might be effective in disrupting the RNA life cycle of a virus.

Helmholtz AI is supporting the project “Learning the language of host-viral protein-RNA interactions: new possibilities for short and long-term intervention” with close to 200,000 euros.

About Helmholtz AI

The Helmholtz AI Cooperation Unit (Helmholtz AI) permanently and strategically drives and coalesces the field of applied Artificial Intelligence within the Helmholtz Association. One essential component is funding research projects that address the topics of applied Machine Learning (ML) and Artificial Intelligence (AI).

More about Helmholtz AI

More about the Helmholtz AI Projects


Image credit (symbol image at top of page): Adobe Stock/Jackie Niam