Computational identification of drug resistance mutations

The emergence of drug resistance is a major challenge in cancer therapy. Besides other mechanisms, mutations at the pharmacological targets of the drugs play a key role. At the moment, these mutations are often only discovered when patients acquire resistance, which causes a time lag in the availability of effective treatment options. Information about likely mutations can have a major impact on cancer patients, because effective drugs could be discovered earlier and would therefore be available as soon as resistance arises. A computational method to identify resistance mutations that have a high probability to emerge upon the treatment with a specific cancer drug was already developed in previous work. However, this method is limited to drugs, for which already information concerning the interaction with the pharmacological target is available and to mutations, which are in close vicinity to the interaction interface. This project aims to extend the applicability of the tool to also be able to investigate drugs, for which data concerning the binding to the pharmacological target is currently lacking. In addition, a new computational method will be developed to analyse mutations, which are located further away from the interaction site and which confer resistance via a different molecular mechanism. This mechanism is among others also responsible for the effects of cancer-causing mutations and the novel computational method will therefore also be able to analyse these mutations. After the new methods have been comprehensively tested, they will be applied prospectively in close collaboration with wet lab scientists. The investigated proteins are the pharmacological targets of drug candidates, which are currently evaluated in clinical trials for the treatment of multiple cancer types. This project focuses on the development of computational methods to comprehensively analyse all possible mutations affecting the drug target. As a consequence, mutations can be identified, which are likely to arise in the clinic. These methods can therefore have a significant impact on the discovery of novel cancer drugs and, ultimately, on cancer patients.

In the course of the FWF funded Standalone project "Computational identification of drug resistance mutations", computational methods to investigate the effects of protein mutations have been generated and improved. The evaluated effects can for example include the impact of mutations on the binding of the body's own molecules such as small chemical compounds, nucleic acids, or other proteins to pharmacologically relevant targets, or the impact on the interaction with drug molecules. The developed methods can be applied to identify mutations in advance, which may play a role in the onset of cancer, or which can cause resistance to drug therapy. In addition, they can be used to elucidate the molecular mode of action of already known mutations. The gained insights can then support the timely discovery of suitable therapies for patients affected by such mutations.