Over the last few months, we’ve made significant upgrades to our Trusted Research Environment (TRE), the secure computing environment where researchers conduct their analyses. More changes are now in the pipeline.

The upgrades are the result of conversations we’ve had with the research community about how we can improve our TRE. We take the feedback we receive from researchers seriously, and we are always looking at ways of enhancing what we can offer. The resources and packages we’re making available are designed to help research teams conduct their work as efficiently as possible.

Improved access to open-sourced packages 

In recent months, we’ve been working to safely enable access to over 1,000 new and commonly requested open-source R and Python packages within the TRE. We’ve also launched a new service allowing researchers to request access to packages not already available on our platform.

We recognise that open-source packages are vital for research analyses. We want to give researchers freedom to run their research studies as they like – but we must always do this safely to protect participants’ data.

The R and Python package request service strikes the balance by allowing researchers to request new packages for the TRE, provided they meet access and security checks.

If a package is already in the system, researchers can access it instantly. If it’s not, they can raise a ticket with us to request that specific package. If we think a package is not safe for release into the TRE, we’ll explain why. This review process typically takes 2 to 3 days.

We’re currently seeing around 200 successful package downloads from researchers a week, so we know the functionality is already making a difference to researchers’ experience.

Enabling workflows

Another enhancement we’re piloting with a small subset of users is enabling the ingestion and use of workflow languages. Starting with our most requested workflow language, Workflow Description Language (WDL), we hope this will allow for large-scale analytical pipeline development and execution. We will be trialling a second popular language, Nextflow, very soon.

The enhancement is still being fine-tuned, but it’s received positive feedback from a group of test researchers. Once implemented fully, it will cut down the workload of researchers who do computationally heavy activities, saving significant time over the course of a study.

Allowing researchers to bring their own tools and software environments

We know researchers want to be able to bring in their own tools with specific configurations of packages, and software versions. This is to ensure they have everything they need to run and reproduce their analyses and pipelines.

Another capability that we are currently testing with a sub-set of users is the scope for researchers to securely introduce their own containerised software environments to the TRE, giving them greater flexibility to use the tools and packages their research requires. This includes facilitating proprietary or organisation-specific tools and intellectual property they may not want to make available to the wider TRE community.

To make this possible, we have introduced a security scanning process which all researcher-requested containers must pass before being available within the TRE. The process allows researchers to introduce their own software whilst maintaining the security of the research environment.

Over the last few months, we’ve tested the capability, with positive feedback.

We plan to roll this offering out gradually because we need to take care to protect researchers’ intellectual property. We hope to make the functionality available to everyone using the TRE by early 2027.

Taking stock

We are confident that these changes are transforming the TRE into a fast and agile tool. Indeed, we’ve already collected feedback that shows an increase in user satisfaction.

But that’s not all we’re doing to make researchers’ lives easier. These measures are just the first tranche in our long-term roadmap to enhance the TRE.

We’ve already upped the cloud credit we give our research teams, from £50 to £500, to help towards computing costs. Other updates over the coming months will include allowing a rapid ingestion of code stored in GitHub and building an AI code assistant for use within the TRE.

We’ll continue to update the you as we release new functionality, so that you can make the most of Our Future Health’s unique data set.