Building Data Assets in Drug Discovery

Large population-based cohort studies like UK Biobank are crucial for drug discovery, providing the statistical power needed to identify small effects that would otherwise go unnoticed. Data assets can be generated in-house or accessed from external sources, each approach having its own advantages and challenges.

Pharmaceutical research demands structured, quality data, with drug development being heavily regulated. Even in less regulated R&D environments, high-quality data remains vital for strong results and avoiding false positives, particularly in early stages.

The Role of Data Assets in Drug Discovery

Cancer genomics exemplifies the importance of data in drug discovery. By comparing the genomes of cancer tumours to healthy tissue in an individual, researchers can inform targeted treatments and patient stratification, aiming for optimal treatment outcomes.

Statins provided another example of how strong data assets contributed to pharmaceutical research. The Cholesterol Treatment Trialists’ (CTT) Collaboration, conducted meta-analyses of numerous statin trials to identify all effects of statin therapy, both adverse and beneficial. This approach, made possible by extensive data assets, allowed for a detailed evaluation of the drug’s overall impact.

Challenges in Building Data Assets

Building effective data assets comes with several challenges. Data collection, integration, and standardisation require careful consideration. Different data sources and modalities require varying quality control procedures, and batch effects can introduce bias. When combining data from multiple sources, working with raw data allows for effective in-house standardisation, giving researchers maximum control over input and design decisions.

Data security and privacy are critical concerns, particularly when dealing with human subjects. Extensive ethical considerations are required, and participants need assurance that their data is being used safely and for public benefit. This trust is essential to the continued success of large-scale data collection efforts.

Strategies for Building Effective Data Assets

Implementing best practices for collecting, organising, and managing data is essential. This includes robust security measures and the use of appropriately trained personnel. Maintaining the trust of research participants is paramount throughout this process.

Partnerships play a crucial role in building these data assets. While challenging in the private sector due to competing interests, collaboration is more common in academia. Different companies have their own interests, but collaboration is key – the more data, the better.

The UK Biobank Proteomics Project (UKBPPP) is a fantastic example of successful collaboration. Multiple pharmaceutical companies funded proteomic data generation in UK Biobank, each with input into sample selection. They had priority access to analyse the resource before its release to the wider research community. Now available to approved UK Biobank researchers, this project demonstrates how collaboration can benefit both private interests and the broader scientific community.

This approach allows for the creation of larger, more comprehensive data assets than any single organisation could achieve alone. It also promotes the sharing of expertise and resources, potentially accelerating the pace of discovery in drug development. As the complexity and scale of data in drug discovery continue to grow, such efforts are likely to become increasingly important in building and leveraging effective data assets.

To find out more about building data assets in drug discovery and how it can benefit your research or organisation, contact our team here.