Northern Ireland Multimorbidity Research and Discovery (NIMRAD) Consortium

Multimorbidity (multiple long-term conditions) is defined as two or more diseases (usually> 5) occurring simultaneously in an individual (usually increasing with age) necessitating multiple treatments.

Developing appropriate healthcare is one of the greatest long term challenges facing health services, driven by an aging population likely to be living with more chronic diseases and increased treatment needs.

These challenges involve addressing: (a) morbidity interactions, (b) polypharmacy, (c) off-target drug-drug and drug-disease interactions, (d) sub-optimal response to medication, and (e) unsustainable healthcare costs.

The scale and complexity of these challenges necessitate a concerted AI-driven personalized medicine research focus on multimorbidity.

Within Northern Ireland there exists multiple Multimorbidity patient cohorts and associated biobanks which have deep phenotypic data and genotypic data, including whole genome sequence data. However these cohorts exist in discrete unconnected silos.

There also exists a considerable computational and artificial intelligence (AI) capacity and capability in academia, the NHS and industry in Northern Ireland (NI), but again this remains unconnected. The UK Biobank has not recruited from Northern Ireland.

Hence, there is a clear opportunity to expand the cohorts available to research, and compare them with UK Biobank / NIHR BioResource cohorts..

The proposed developmental award will (a) bring the multimorbidity datasets together as a federated resource for advanced computational Artificial Intllegence analyses and (b) create a multimorbidity network of expertise to develop future research that will: 1) comprehensively identify the disease combinations and proportions in our federated NI Datasets for comparison and contrast with the UK Biobank /NIHR BioResource. 2) build diagnostic tests, determining whether a patient is multimorbid and which combinations of disease they have or are at risk of developing by extracting the clinical, physiological and genetic data of patients in our federated NI Datasets with significant disease combinations and use machine learning methods.

This will include the use of clinical, geographical and socioeconomic deprivation measures and details of the interactions with healthcare obtained from electronic healthcare records (ECR) to identify risk factors and their relative significance for developing multimorbidity and patient trajectories for the development of long-term conditions. 3) explore the cellular and tissue level mechanisms that are responsible for determining whether a patient is likely to develop a single disease or multimorbidity.

The protein, genes and features that distinguish between morbidities will lead such exploration.

Proteins and genes that do not distinguish between morbidities, but do distinguish between health and disease will be used to explore the underlying mechanisms driving multiple diseases. 4) suggest therapies that maximally drive disease regression while minimising off-target effects, by identifying drugs that target proteins and genes involved in these cellular and tissue level mechanisms and using computational optimisation to identify optimal dosages and combinations.

This proposal will (a) develop integrated access to new cohorts for research and (b) undertake foundation work for future studies of i) how multimorbidities cluster; ii) the associated physiological, environmental and sociological risk factors; iii) the underlying mechanisms driving multimorbidity; iv) diagnostic blood tests suitable in clinical care; v) therapeutic strategies for treating patients with multimorbidity.