The University of Oxford works across a broad range of research centres and consequently handles data in various different forms from various sources. Data including from DNA sequencing, transcriptomics, proteomics, and clinical metadata all need to be integrated to enhance data accessibility and analysis.
Over the pandemic, the University of Oxford undertook a multimodal project which investigated data from over 100 patients with different diseases including COVID, influenza, and sepsis, as well as healthy controls. 200 researchers across the university came together to look at data from clinics, transcriptomics, cytometry, Olink, and proteomics. This was integrated into a large dataset which was released to the research community.
Stephen Taylor’s team have been working on data visualisation and analytics to improve access to these complex datasets and drive learnings from them. This led to a project called Multi Locus View, which saw the development of a web-based tool for the analysis of genomic data. Further tools were integrated into the project, including a machine learning algorithm called LanceOtron, trained on sequencing data.
The project expanded through the use of a Hyperion system into Multidimensional View (MDV) which could handle multiple heterogeneous datasets. Taylor’s team at the Centre for Human Genetics has since expanded to work on this project.
They start with data preparation: capturing, naming, and pseudo-anonymising it to comply with information governance. Then the ingestion stage, where the data is harmonised, standardised, and registered. Finally, the analysis and visualisation stage where the datasets are queried. This ensures that any published data is readable and understandable so that new learnings can be produced.
Furthermore, the Spatial Omics Oxford Pipeline (SpOOx) is a newly developed pipeline designed for spatial statistics and data integration. It allows for detailed analysis of immune responses and cellular interactions by integrating various data sets, including imaging data, and applying sophisticated spatial statistics methods. This pipeline is part of Oxford's broader efforts to enhance data visualisation and analysis capabilities.
