Machine learning prediction of neurodegenerative therapy response using molecular biomarkers

Apply for a PhD studentship at the Neuroscience Institute.

About the project

The failure rate for new drugs entering clinical trials is in excess of 90 per cent, with more than a quarter of drugs failing due to lack of efficiency. In order for there to be good efficacy in clinical trials, there should be good evidence of efficacy in preclinical drug tests.

In particular, for testing new therapeutics, it has been difficult to reproducibility determine drug response in cell lines, which lack ways to quantify uncertainties in measuring in vitro dose-response.

Gaussian Processes (GP) is a machine learning technique that we have successfully used to estimate uncertainty in noisy gene expression datasets and predict emotions from dynamic facial features. In this project, we will take this idea one step further by using GPs to model uncertainty for in vitro dose-response together with molecular information and then predict probabilities of response for new treatments using molecular features.

We hypothesize that the probability of response for new therapeutics targeting neurodegenerative disease can be predicted using genomic, cellular and clinical features.


  • Dr Mauricio A Álvarez
  • Dr Dennis Wang

Entry requirements

A first or upper second class honors degree or significant research experience.


Dr Mauricio A Álvarez
+44(0) 114 222 1902

How to apply

You can apply using our Postgraduate Online Application Form. State the prospective main supervisor on the online applicaiton form and select (BMS) as the department. For mone on applying see our applying essentials page.

Apply now

PhD scholarships

There are lots of different types of scholarships available, from full awards covering fees, living expenses and research costs to partial fee scholarships.

Apply now for a scholarship