PhD Studentship Opportunity: Multimodal analysis of plasma metabolomics and plasma proteomic data with respect to chronic diseases in UK Biobank volunteers

UK Biobank recruited 503,000 volunteers aged 40-69 in the years 2006-10. These volunteers completed (and continue to complete further) detailed questionnaires on their health, lifestyle and preferences (e.g. food). At the same time, physical measurement data were collected and a number of biological samples were obtained. All of the participants have been genotyped and they will all have their entire DNA sequenced by 2022. Biochemical analysis of plasma and other samples will be an ongoing process over the coming years. At present there is extensive biochemical data on the blood from volunteers. This adds to longitudinal follow-up data from web-based questionnaires and access to primary health care records of UKB volunteers. UK Biobank is seeking to develop its data in metabolomics and proteomics: information on the levels of ~200 metabolites and ~2000 proteins in blood is currently being acquired. This huge amount of clinical, genomic, genotypic, proteomic, metabolome, questionnaire and latterly COVID-19 related data, is a treasure trove from which we can extract data on disease risk, environment and health, biomarkers for wellness, diet and disease, and many other areas.

Peripheral blood offers major opportunities to identify and quantify biomarkers of wellness and disease. UK Biobank has measured the concentration of hundreds of metabolites and thousands of proteins in the same UK Biobank participants’ samples. It is clear that there can be a wide variation in the levels of a specific biomarker within the human population. Presently a reference range (concentration values that include upper and lower limits of an analyte based on a group of apparently healthy people) is used to define the value of a biomarker used in a clinical test (for a single analyte such as C Reactive Protein). With the advent of AI and machine learning approaches to defining markers of disease risk, prognosis and diagnosis, a knowledge of concentration range values for as many proteins and metabolites as possible would benefit us to develop tools to improve clinical outcomes. To achieve this a determination on how concentration values vary with respect to age, sex, disease and comorbidities is required. Furthermore, analysis of the associations between protein and metabolite entities in the blood will enable us to understand whether constituents of a specific biochemical pathway are all modulated in the same fashion. In this project we will employ bioinformatics and machine learning approaches, applied to the extremely rich UK Biobank dataset to define reference ranges for plasma proteins and metabolites. We will determine correlations between specific metabolites and proteins and chronic clinical conditions such as respiratory disease and arthritis; as well as determine if specific pathways are associated with clinical conditions by looking for significant changes in proteins/metabolites in the same biochemical/physiological pathway.

The student will be taught advanced informatic, machine learning and artificial intelligence techniques in order to undertake the work described above.

Entry Requirements

Upper second class honours degree, for example in biological sciences or computer sciences, or the equivalent, which can include relevant work experience.

IELTS minimum 6.5 overall with 6.0 in Writing, or equivalent. 

How To Apply

If you think you are interested in the post and want an informal discussion then please submit a curriculum vitae and short letter to [Email Address Removed].  Formal applications should be submitted via the Biosciences and Medicine PhD programme page on the "Apply" tab. Please clearly state the studentship title and supervisor on your application.

Start Date

October 2022