For several decades, progress on effective therapeutics for dementia and neurodegenerative conditions has been frustratingly slow. Although the first generation of Alzheimer’s drugs have now arrived in the form of Leqembi and Kisunla, for the most part the promise signalled by preclinical breakthroughs hasn’t materialised into the much-needed, effective treatments sought across neurodegeneration. There are several reasons for this but crucially limited access to human models and data has been at the root of poor translation in the neurodegenerative field.
One approach that is already transforming the medical field and holds much promise for dementia and neurodegeneration is human data mining – the systematic analysis of large-scale health databases. UK Biobank is a repository of records and samples from over half a million UK residents, including environmental and lifestyle information, medical records, genetic and other biomarker data. Over the past few years, findings obtained from mining these records have dominated the science headlines, including UK DRI-led studies revealing that sleep disorders put people at greater risk of dementia and that smartwatches could detect Parkinson’s up to seven years before hallmark symptoms appear.
By 2027, it is estimated around 40,000 UK Biobank participants will sadly be diagnosed with dementia. However, alongside other health databases including the recently announced health data research service to facilitate fast, secure access to anonymised NHS records, these resources can provide unprecedented insight into disease and speed up the delivery of new therapeutics. For progressive, age-related conditions such as Alzheimer’s longitudinal studies like Our Future Health are vital to unpick early life risk factors and underlying disease causes.
Although these population-scale studies are now unlocking new insights into neurodegenerative conditions, these resources can be leveraged further with better integration between cutting-edge, fundamental biological research and clinical and genetic datasets. We are now in an era where laboratory discoveries made in cells and observed under microscopes can be validated quickly at a large-scale in human populations, rather than remaining isolated before clinical testing. This integration has the potential to transform our search for treatments by dramatically reducing the time and expense of drug development. Population data reveals which biological pathways are truly relevant in humans, identifies patient subgroups most likely to benefit from specific therapies, and helps prioritise which laboratory discoveries merit costly clinical trials.
As a data analyst, I am excited to collaborate with fundamental biologists and clinicians, jointly recognising the importance of integrating hypothesis-driven and data-driven approaches to ensure a bidirectional flow of insights that bridges the mechanistic understanding of dementia with clinical relevance.
Group Leader
Recent collaborations demonstrate the power of this integrated approach across different aspects of neurodegeneration research. Dr Bhuvaneish Selveraj (UK DRI at Edinburgh) identified a novel molecular target for motor neuron disease using patient stem cells and tissues, then partnered with Prof Valentina Escott-Price (UK DRI at Cardiff) to validate this discovery using genetic data from UK Biobank's half million participants - providing statistical power impossible with traditional studies and guiding drug development strategies. Similarly, a team led by Prof John Hardy (UK DRI at UCL) and Prof Escott-Price uncovered previously hidden genetic interactions in Alzheimer's disease by focusing on individuals with the highest genetic risk, revealing new disease mechanisms in the DAB1-RELN pathway that conventional analysis had missed.
These discoveries extend beyond laboratory validation to direct clinical impact. Working with Prof Zameel Cader (University of Oxford), the Cardiff team used 20 years of Welsh health records from the SAIL database to prove that sustained blood pressure control independently reduces dementia risk - transforming a biological hypothesis into evidence-based support for active blood pressure management as a prevention strategy. Together, these examples illustrate how human data mining accelerates discovery timelines, reveals insights invisible in smaller studies, and ensures research translates into actionable clinical interventions.
The UK DRI Human Data Mining Hub
The lab of Prof Valentina Escott-Price (UK DRI at Cardiff) pioneers the use of population-level data to uncover insight for neurodegenerative conditions. Collaborating closely with fundamental biologists she has recently established the UK DRI Human Data Mining Hub alongside Dr Emily Simmonds. The Hub provides both technical infrastructure and collaborative expertise to bridge fundamental biology with population-scale insights, facilitating this integration for neurodegeneration researchers across the Institute.
The Hub's impact is now expanding internationally through a new strategic partnership with South Kazakhstan Medical Academy. This three-way collaboration agreement with Cardiff University will establish computational facilities and train local researchers to analyse electronic health records, representing an international expansion of the UK DRI's human data mining capabilities to new populations and disease contexts. Another partnership recently announced between the UK DRI and Centre for Brain Research in Bengaluru, Indian Institute of Science, will also benefit from the Hub’s expertise for establishing large-scale healthy ageing studies in India. By working with international partners to develop these analytical capabilities, the Hub is helping to unlock the potential of human data mining for neurodegeneration research on a global scale, ensuring that discoveries made through population-scale analysis can benefit patients worldwide.
For UK DRI researchers looking to find out more about the Hub, please visit the UK DRI Human Data Mining Hub page on the Portal.
