DARE UK Sprint Project: Trusted Research Environment and Enclave for Hosting Open Original Science Exploration
Trusted Research Environments (TREs) are secure computing environments providing secure access to sensitive data for research purposes. HDR UK has a set of resources and guides explaining TREs in more detail in terms of health data research, although TREs are not exclusive to health data.
TREEHOOSE is an open-source platform for deploying TREs on Amazon Web Services (AWS). It will include open-source tooling to streamline building and operating TREs on public cloud infrastructure whilst maintaining security and trust.
The recent Goldacre Review ("Better, Broader, Safer: Using Health Data for Research and Analysis") highlighted the need for standardisation across TREs, ideally through the use of open-source infrastructure.
TREEHOOSE is under active development. It is suitable for anyone interested in deploying a trusted research environment on AWS. Currently it has good support for launching customised Windows Desktops, and limited support for Linux workspaces with SSH access. All access is managed through a TRE web interface which prevents unauthorised egress of confidential data.
Features include automated backups for researcher workspaces, secure egress requiring approvals from data governors or other authorised personnel, and budget alerts to help manage spending.
You can deploy a TRE on your own following the TREEHOOSE documentation, but due
to the active development we strongly encourage you to get in touch with us
first, either by
opening a GitHub issue on this repository
or by emailing [email protected]
.
We can help demonstrate features of the platform and see how they match your requirements, and discuss future enhancements.
TREEHOOSE was originally developed for use with confidential healthcare data such as patient electronic health records, but is designed to be used and customised for all research and analysis disciplines which require access to sensitive data.
The documentation is divided into several sections:
The main purpose of this repository is to continue evolving TREEHOOSE, making it faster and easier to use. Development of TREEHOOSE happens in the open on GitHub, and we are grateful to the community for contributing bugfixes and improvements. Read below to learn how you can take part in improving TREEHOOSE.
TREEHOOSE has adopted a Code of Conduct that we expect project participants to adhere to. Please read the full text so that you can understand what actions will and will not be tolerated.
Read our contributing guide to learn about our development process, how to propose bugfixes and improvements, and how to integrate your changes in this repository.
This project is licensed under the Apache-2.0 License.
This work was funded by UK Research & Innovation Grant Number MC_PC_21032 as part of Phase 1 of the DARE UK (Data and Analytics Research Environments UK) programme, delivered in partnership with HDR UK and ADRUK.