Approved Research

Treatment selection for depression based on genetic predictors

Lay summary

The aim of this research is to enhance the treatment of depressed patients. In specific it is intended to find better mechanisms to select the most promising treatment for an individual patient. For this the expected success of on average similar effective treatments like psychological treatment and antidepressants is compared.

Previous research has shown that genes, age, sex, and income among other things influence the risk to suffer from depression. Additionally, novel research has revealed that similar factors also influence the treatment success. The large dataset of the UK Biobank contains many genetic information about patients. Additionally, the dataset includes information about the diagnosis of depression and indirect information about the treatment success. Novel machine learning and artificial intelligence methods can find complex patterns based on large datasets to predict the individual treatment outcome.

The project duration is scheduled for 36 months. The first 12 months the project will focus on data analysis. The second 12 months will be dedicated to identifying the patterns in the data with novel machine learning methods. The last 12 months will be used to summarise and publish the results.

The research project can in the case of success influence and improve the public health in two ways. First, for patients suffering from depression the chance to be treated successfully in the first attempt would increase. This would reduce a large amount of individual suffering. Second, for the public and health care payers the overall treatment costs for depression will be reduced. Especially the high costs for intensive impatient treatment could be avoided in several cases. Additionally, for the public the high costs of sick days due to depression could be reduced.