This project centres around better understanding obesity, a disease whose prevalence continues to rise worldwide despite efforts to tackle the epidemic. Trends of increasing overweight and obesity are important because they are potent risk factors for many other diseases, prompting clinicians to call for action (https://doi.org/10.1016/S2213-8587(22)00317-5). Whilst there is evidence that adiposity, measured by body mass index (BMI), causally influences a range of health outcomes, there is little understanding of the biological mechanisms driving BMI effects (https://doi.org/10.1002/oby.21554). Furthermore, there is increasing recognition of the limitations of BMI as a measure, including new clinical recommendations relating to definition and diagnosis of obesity (https://doi.org/10.1016/S2213-8587(24)00316-4). Metabolomics is the large-scale study of metabolites (small molecule substrates, intermediates, and products of cell metabolism). Current approaches in the field of metabolomics enable the measurement of hundreds of metabolites from low-volume samples. These data contain information relevant to a wide range of health conditions and can help understand the complex link between risk factors and downstream health outcomes (https://doi.org/10.1136/bmjmed-2023-000787). We have conducted a series of metabolome-wide association studies (MWAS) for BMI using different study designs (https://doi.org/10.1007/s00125-023-06019-x, https://doi.org/10.1002/oby.23441). However, this univariate MWAS approach of identifying individual metabolites associated with a given exposure or outcome without any consideration of the interrelationships between metabolites is likely a sub-optimal approach. We propose that by using a network analysis approach applied to metabolomics data we can identify biological signals relevant to BMI and its impact on downstream health outcomes.
Aim: To integrate data from various study designs and to use network analysis and/or machine learning methodologies, to elucidate the molecular footprint of body mass index.
Methods: This project will make use of metabolomics datasets from two complementary commercial platforms. Firstly, Metabolon’s mass spectroscopy-based platform that delivers high quality semi-quantitative data for more than 1400 metabolites from a single sample, providing excellent coverage across the full spectrum of molecules found in the circulation. Secondly, Nightingale Health’s proton nuclear magnetic resonance (NMR) spectroscopy platform that provides a detailed quantification of circulating plasma lipoprotein lipids and a selection of amino acids and carbohydrates. Studies include (but are not limited to):
1) The Diabetes Remission Clinical Trial (DiRECT)
2) The By-Band-Sleeve Trial
3) The Avon Longitudinal Study of Parents and Children (ALSPAC) including data from age 7 to age 30.
Objective 1: Characterise the properties of metabolites and metabolic profiles relating to BMI within studies
To integrate metabolomics data across multiple study designs, there needs to be a good understanding of the properties of the metabolites and their interrelationships and how these vary within and between studies. The aim in this part of the work is to develop an analytical pipeline to characterise metabolites in a multivariate framework. Work might include, for example, an assessment of the extent to which relationships (profiles) between metabolites are consistent and reproducible across different datasets. Methods to compare Gaussian graphical models (https://doi.org/10.1093/ije/dyy119) could provide a starting point.
Objective 2: Apply network analysis and/or machine learning methods to integrate and to compare metabolomic signatures of BMI across interventions and population-based analyses
Here profiles or networks of metabolites and their relationship with BMI will be compared with a view to identifying more biologically meaningful patterns of association than those
from univariate statistics. Methods developed for multi-omics data integration will be reviewed and potentially adapted to the current application. Two reviews of integration methods (focused on metabolomics) have been published providing a useful starting point (https://doi.org/10.1016/j.aca.2020.10.038, https://doi.org/10.3390/metabo9040076). The student will also consider adaptation of cutting-edge methods from AI and Machine Learning such as Graphical Neural Networks (https://arxiv.org/abs/2405.19230; ICLR 2025) to omics. Taking forward the best identified method, the second step will be to modify and test the efficacy of the method when applied to our research question and datasets.
Based on their critical review, the student will select their method of choice or develop novel methodologies uniquely appropriate to multi-study single omics integration.
Objective 3: Explore the relevance of findings to BMI-associated health outcomes
Taking the learning from Objective (2), the student will consider how the metabolomic profile of BMI relates to relevant health outcomes utilising additional cohort data, for example, UK Biobank. Initially the focus will be on endometrial cancer where findings will be considered in the context of results from alternative approaches applied by the supervisory team, e.g. Mendelian randomization studies (https://doi.org/10.1101/2024.04.18.24305987). The student will then be able to choose other disease outcomes according to their interests.
Funding
These studentships are funded for four years full time through GW4BioMed2 MRC Doctoral Training Partnership. Part time study may also be available. The studentships consist of tuition fees and a stipend matching UK Research Council National Minimum (£20, 780 p.a. for 2025/26, updated each year). Additional research training and support funding of up to £5,000 per annum is also available.
Eligibility
Please see the GW4 website for information about eligibility and entry requirements, including specific information for international applicants: GW4 BioMed2 Student FAQs.
How to Apply
A list of all the projects and how to apply is available on the DTP’s website: gw4biomed.ac.uk. You may apply for up to 2 projects and submit one application per candidate only.
Please complete an application to the GW4 BioMed2 MRC DTP for an ‘offer of funding’. If successful, you will also need to make an application for an 'offer to study' to your chosen institution.
Please complete the online application form linked from the DTP’s website by 5.00pm on 20th October 2025. Please note that we may close the application process before the stated deadline if an unprecedented number of applications are received– check the DTP’s website for details and updates. If you are shortlisted for interview, you will be notified from 23rd December 2025. Interviews will be held virtually on 27th and 28th January 2026. Studentships will start on 1st October 2026.
Further Information
Informal enquiries: GW4BioMed@cardiff.ac.uk
Project-related queries: svetlana.mastitskaya@bristol.ac.uk
