Novel biologically informed approaches to understanding and treating depression and related mental health conditions
Approved Research ID: 92794
Approval date: November 29th 2022
This project aims to use the latest advances in genomics, artificial intelligence, and brain imaging to solve a number of outstanding questions and problems pertaining to the treatment and understanding of depression. Those are:
1) There has been no major innovation in antidepressant drugs in the past 30 years.
Recent research in other diseases suggests that there might be ways to find new drug targets based on genetics. Because of the genetic element of mental illness, it is possible to use large-scale genetic data sets to find new ways to identify and develop drugs. This can be done using a combination of two methods called Mendelian Randomisation and Machine Learning. In essence, we almost simulate a clinical trial from existing genetic studies in UKBiobank. Finding these targets using this method is a novel way and will provide new insights that can be utilised in drug development.
2) We know that depression and other mental health and behavioural conditions like gambling conditions are linked. However, there is no way to know if one causes the other. Our approach will use the novel method of Mendelian Randomisation to stimulate a clinical trial that definitively find or refute links between depression and gambling.
3) There is no streamlined approach to predicting response to antidepressants in patients with depression, even though almost half of the patients will not respond to the first treatment approach. Some genetic tests are available but on their own they don't predict treatment response accurately enough. Further, brain imaging studies have so far been relegated to research domains. Our approach is to create an AI algorithm that combines brain imaging and genetic markers to predict who will respond to certain antidepressants and who might need to change the dose and or medication type. This will be an important step in delivering the promise of personalised medicine in psychiatry.
4) Depression impacts people in distinct ways, and the reasons for this are not well understood. For example, it is currently unknown why, despite experiencing symptoms of similar severity, one person with depression may be able to maintain full time employment while another may find work extremely challenging. Our approach will use AI to help individuals predict how their experience with depression will pan out, and how anticipating certain support needs could help to reduce depression-related disability and its personal, social and economic costs.