---
tags: satre
---
# SATRE online Collaboration Cafe | 18 May 2023

:::info
If you are new to HackMD, please see this document for a short guide (right click, open in a new window): [https://hackmd.io/@turingway/hackmd-guide](https://hackmd.io/@turingway/hackmd-guide).*

:::

:wave: We're delighted to have you here. All are welcome - no sign-ups needed. (We do recommend bringing your actual tea, coffee, water, etc. to this cafe! :coffee: :sparkles: :cake:

**When?**
18 May 2023, 15:00 - 16:00 Europe/London ([see in your time zone](https://arewemeetingyet.com/London/2023-05-18/15:00))
[Schedule for future calls](https://hackmd.io/GI5EZQouTYuGhkYBSBYHmQ)

:::info
**Zoom Information***:
* https://turing-uk.zoom.us/j/99226214703?pwd=VWRaWXhacitRS25KODRLOXN3bkI2UT09&from=addon
* Meeting ID: 992 2621 4703, Passcode: 291666

:::

**What?**
*[SATRE (Standardised Architecture for Trusted Research Environments)](https://medium.com/satre) is a DARE UK Driver Project working to standardise access to secure data in trusted research environments. It includes University of Dundee, Alan Turing Institute, UCL, Ulster University, Research Data Scotland.*.
[Collaboration Cafes](https://the-turing-way.netlify.app/community-handbook/coworking/coworking-collabcafe.html) are **online collaboration and coworking calls** for anyone interested in learning about, discussing and contributing to the project.

**Who?**
***Anyone** interested in our project & the TRE space is welcome to join this call. No prior sign-ups needed!*

***All questions, comments, and recommendations are welcome!***

---

## Check-in and icebreaker

**Name + Pronouns + Introductions + Location + x + an emoji ([emoji cheatsheet](https://github.com/ikatyang/emoji-cheat-sheet/blob/master/README.md))**
*(Remember that this is a public document. You can use a pseudonym if you'd prefer.)*

* Arron Lacey (he/him) Senior Community Manager @ The Alan Turing Institute :tophat:
* Cian O'Donovan (he/him) social scientist in UCL and researcher on SATRE project 🧶
* Antony Chuter (He/Him) Lay Patient/Public member of Alleviate, and SATRE, Uckfield in Sussex
* Wataru Suzuki (he/him) NHSE Architect on Secure Data Environment(SDE)
* David Sarmiento (he/him) + Research Project Manager at Turing + Edinburgh
* Hari Sood (he/him) + RAM at Turing Institute + London + :surfer:
* Michal Rosiek (he/him) TRE Lead at KCL (King's College London)
* Pete Barnsley Francis Crick
* Madalyn Hardaker (she/her) Head of e-Research Data Governance, King's College London, :sunny:
* Chris Cole (he/him) - PI of SATRE project & Senior Lecturer, School of Medicine, University of Dundee
* Ed Chalstrey - Turing Data Safe Haven dev
* Will Crocombe - freelance IS/IG
* Matt Craddock (he/him) + Research Software Engineer - Alan Turing Institute + Lincoln + :coffee:
* James Grant - Research Solutions Architect - Amazon Web Services
* Jim Madge + he/him + Senior Research Software Engineer @ Turing - Exeter + 🍵
* Tim Machin - UCL, Senior Enterprise Architect (Research) working with SATRE
* Matt Penn - KCL Head of Scientific Computing Infra (he/him)
* Vivek Iyer - Sanger Human Genetics
* Adam Huffman - Big Data Institute, University of Oxford
* Adam Booth (he/him) - Our Future Health
* Andy Shepherd - University of Manchester / NERC Digital Solutions
* Claire MacDonald - Manchester University Hospitals NHS FT (MFT) - Head of Data Science
* Sarah Waters (she/her) - University of Oxford - Information Security Governance Risk and Compliance Officer
* Keiran Raine (he/him) - Eastern AHSN - Head of Service, Clinical Computational Research Informatics
* Lucy Cheesman - University of Sheffield - Research Data Services Manager

## Breakout rooms: Topic proposals and notes

*While no sign-ups are required to attend Collaboration Cafe, if you have an idea for a topic you'd like to discuss in a breakout room, please add it below and put your name next to it. A good place to start to find discussion topics are on the [SATRE Specification Repo Issue Board](https://github.com/sa-tre/satre-specification/issues)*. If you're breakout room is discussing an Issue, we encourage you to write up any thoughts directly in the Issue thread.

* Main room: Survey Share-out (HS,JG)
* Breakouts Rooms:
* Intro to project and GitHub repository (HS on Tuesdays, AL on Thursdays)
* User engagement (COD)
* [name=Ed] Is this separate from the survey responses?
* IG: https://github.com/sa-tre/satre-specification/issues/69
* Tiering: https://github.com/sa-tre/satre-specification/issues/68
* TRE purpose and how overlaps and drives purpose and spec
*
*

### Break out notes

#### IG

Chair: Hari

- ... initial discussion not captured
- [name=Hari Sood] Focus on content of issue (68)
- proposal - supporting different sensitivities is a requirement, preset configurations is an optional feature
- [name=Tim Machin] Mix of 'requirements' and 'ways to implement'. Tiering is an example of the later. The meaning of tiering is quite different for models where work is done in a ephemeral instance vs. a large persistent data store. What would be useful in the specification is example of Tiers, requirements and suggestions of how to achieve this rather than _specific_ requirements
- [name=Wataru Suzuki] Tiers are likely to be very domain and context specific (risk appetite). Consider modifying data to meet research requirements rather than meeting controls to data.
- [name=Hari Sood] Feels like feeling of risk appetite and difference of appraoches across domains is being captured. Supporting different levels of sensitivity is fairly, but not universally popular, configurations less so.
- [name=Andy Shepherd] May be necessary to distinguish between privacy and security, what does sensitivity mean?

#### TRE Purpose

Chair: Chris

- how can make collaborative research speedy, efficient, low cost (in TREs)?
- Requires governance, infrastructure.
- Question. Is the purpose of TREs provide secure access to data or to enable research?
- TRE is an administrative system for IG, legal, Data sharing agreements etc.
- Data is not the difficult issue.

### Link List ✨
* Github issues: https://github.com/sa-tre/satre-specification/issues (50-69 are survey MCQ answers for discussion)
* The spec these will go into: https://satre-specification.readthedocs.io/en/latest/
* Glossary issue for anyone who wants to work on this: https://github.com/sa-tre/satre-team/issues/15
*

#### User engagement

People: Cian, David, Sarah, Michal, Claire

Assistants are finding challenges in engaging with users and to work with them in setting up systems.

SATRE project will host two user engagement sessions next week:

* 24th May 11.30-13.00 BST & 26th May 14.00-15.30 BST
* [Register here](https://forms.office.com/Pages/ResponsePage.aspx?id=p_SVQ1XklU-Knx-672OE-UJvD84dTx5NvvzlldEbytBUMUIzUFJPN01HSDIwNDVFNDlEU1MyTTZEQy4u)
* Agenda
* Introduction to TREs
* The SATRE project- activities and goals
* An open-source reference architecture
* Group discussion of users experiences
* How to engage with the reference architecture as a user

Is it worth categorising users across types? create user personas?
* Technical, open to innovation users vs "pointed by the finger" who want to be guided and stay with what they know
* Some may assume this (a system) is not for them, yet we need to engage with them
* Provide plenty of materials in advance

## Agenda

| Time | Activity |
| ---- | -------- |
| **12.05 - 12.15** | Intro to SATRE and Survey Share-out (HS,JG) |
| **12.15 - 12.25** | (1st breakout session) |
| **12.25 - 12.30** | (☕️ Break) |
| **12.30 - 12.55** | (2nd breakout session) |
| 5 mins | :wave: Reflections and close |

### Notes

* Q: Will we look for patterns/clusters in the survey responses?
* This is do-able for Likert responses
* Q: Even though survey responses came from a diverse population, could there be any bias i.e. lots of software engineers that result in a skewed specification. How do we deal with this?
* Community consensus will be important to address this and define what we might be missing as the specfication takes shape.
* Some responses were from an organization rather than indivduals - perhaps they should carry more weight?

### Different goals for TREs / how to decide what should be required vs optional (Room 4)

#### People
- Pete Barnsley, Francis Crick Institute
- Neil Prockter, LSE
- Chris Cole, University of Dundee
- Martin O'Reilly, The Alan Turing Institute

#### Purpose of TRE
- Support speedy, cost-effective, high quality research
- Best problems usually require most collaboration
- Collaboration requires ethical framework, legal framework, financial framework and associated collaboration infrastructure
- At Crick focus is on supporting collaboration, not using TRE just to allow access to certain datasets.
- At Turing similar, goal is to maximally support researchers in productive research with widest access to compute, analysis and collaboration tools, while keeping sensitive data secure enough.
- A TRE isn't just the underlying technology (or even primarily). It's the mechanism to support the chosen IG process, which will differ across different organisations, funders, ethics processes, data providers.
- Goal of SATRE is to align on common approaches where possible and identify commonalities and differences in IG and processes and technical mechanisms to support that.

### Feedback at the end of the call

* We will work on a way of capturing contributions for those not so familiar with GitHub
* HackMD is a barrier to entry
* Google Docs is more familiar
* But might struggle with lots of usage
*
* We could do with more time in the breakout rooms next time
* And we could provide scribes to capture HackMD notes
* Can we have e.g. a sprint version where every 2 weeks we release a version that people can comment on
* Can we have 10/15 minutes at the beginning to bring people up to speed with HackMD
* Or instructions we record and share out beforehand
* We can do GH training for just engaging in the issues
* Is it worth starting with specific issues that we can get peoples thoughts on, e.g. bring specific issues we want community input on
* Then have scribes/note taking for specific issues
* How to convert issues to PRs - how to encourage this? Are insisting on using forks best?

----

### :link: Useful links

* Medium blog: https://medium.com/satre
* Specification repository: https://github.com/sa-tre/satre-specification
* Contact form: https://forms.office.com/e/FuFyNGx3hw
* RSE TRE community mailing list: https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?SUBED1=RSE-TRE-COMM&A=1

### Code of conduct

* [Take a moment to read this](https://github.com/sa-tre/satre-specification/blob/main/CODE_OF_CONDUCT.md)

---

:::info
**Archives of previous Collaboration Cafe calls are hosted in this schedule note: https://hackmd.io/GI5EZQouTYuGhkYBSBYHmQ**
:::