We are developing bioinformatical models for predicting viral resistance based on viral genotype and additional clinical parameters and offering them freely via the internet for clinical research and patient treatment.
The resistance of a virus is coded in its genome. Sometimes, very few individual mutations are crucial for viral resistance, sometimes a larger number of mutations bring resistance about. In general, it is hard to decide upon resistance and select effective therapies by manual inspection of the genome. Laboratory test that can inform about resistance are difficult, expensive and take a long time. Thus we have resorted to computer assistance for this purpose. We mine large collections of virologic and clinical data with bioinformatics methods to afford predictions of the level of resistance of a virus to an individual drug and of the estimated effectiveness of a therapy comprising a collection of drugs. The viral genome is taken from the patient to be treated and the suggested resistance levels and drug combinations are specific to that patient. Thus we have a scenario of individualized therapy.
Use of our bioinformatics methods is available for free via the geno2pheno Server on the internet. The Laboratory for Resistance Analysis at the Institute of Virology at Cologne University is the main virologic partner in this initiative. Until 2018, our main bioinformatic partner was the Department for Computational Biology and Applied Algorithms at the Max Planck Institute for Informatics in Saarbrücken, Germany, which was directed by Prof. Thomas Lengauer. Due to his retirement in 2018 that department was dissolved. Prof. Lengauer is now affiliated with us to continue this research.
The respective research has been carried out in the context of several national and European collaborations, notably the Arevir and Euresist Consortia.