Lab Meeting on 10 October 2019

Posted on by Kirstie Whitaker

The week’s lab meeting fell on a Data Science for Mental Health Interest Group meeting and at the same time as a Turing Way development sprint, so the updates were asyncronous.

It was also World Mental Health Day so we collected together some recommendations to take care of yourself and your mental wellbeing.

If you didn’t receive the Turing Way newsletter you can read it (and sign up for it) at https://tinyletter.com/TuringWay/archive ✨

Celebrations and cool things to share

Kirstie is really excited to be hosting a Turing Way sprint this week! They’re discussing how to extend the project to include ethics, scoping and designing a data science project, communicating your results (outreach, visualisation, how to write a paper etc) and guides for effective and inclusive collaboration. 🙀🚀💫🙌

Malvika and Kirstie will be co-working at MozFest House on Thursday 24 October between 10am and 4pm. It would be lovely to see as many lab members there as possible (at the regular lab meeting time, 3pm) for a general update on the Turing Way project.

Kirstie was also excited to share that the Open as in Book discussion this month is right at the end of the sprint - at the time she’ll need it most - and on one of her absolute favourite books: Emergent Strategy by adrienne maree brown. You can read more about it (without buying the book) at https://praxis.fortelabs.co/emergent-strategy-organizing-for-social-justice. She shared the principles from the book:

  • Small is good, small is all (The large is a reflection of the small)
  • Change is constant (Be like water)
  • There is always enough time for the right work. There is a conversation in the room that only these people at this moment can have. Find it.
  • Never a failure, always a lesson
  • Trust the People (If you trust the people, they become trustworthy)
  • Move at the speed of trust
  • Focus on critical connections more than critical mass—build the resilience by building the relationships
  • Less prep, more presence
  • What you pay attention to grows

Konrad met this week with someone whose role at UCL is to facillitate the translation of basic science to clinical practice. He found it eye-opening to think about all the parts of a translational project that are separate to any of the basic science: needing a project manager, business plan, considering intellectual property, additional funding, and managing clinical trials.

Patricia One of the SSI Fellows currently runs a survey on remote collaboration - it would be great if lab members could complete the form.

Georgia has been running and analysing focus groups with autistic people and developers. She’s found it really interesting to see how ideas are being sparked. Its tough not to inadvertantly lead the conversation!

Yini shared a book she was reading: “Madness and Civilization: A History of Insanity in the Age of Reason”(https://www.amazon.co.uk/Madness-Civilization-M-Foucault/dp/067972110X) by French philospher Michel Foucault. The book changed a lot of people’s perceptions of madness and also changed the prejudices about some specific groups.

Ang shared a paper recently published in eLife: “Science Forum: Ten common statistical mistakes to watch out for when writing or reviewing a manuscript” (doi:10.7554/eLife.48175).

Sarah was interviewed for her friend’s blog about her decision to leave academia. (“Spoiler alert: best decision ever!") You can read it at https://exploringnewworlds.home.blog/2019/10/10/interview-with-sarah-gibson-research-data-scientist-the-alan-turing-institute.

Isla was awarded a distinction for her masters dissertation! Congratulations!! 🎉🎉🎉 from all the lab 🏆

Elizabeth has been working on her JavaScript to try and help Jupyter Book… which has in in-practice meant working on her website builder. She recommended Vue.js if you haven’t tried it yet 🌺

Questions we’re thinking about

Kirstie wants to know your thoughts on what would motivate PhD students, PIs and business team members to contribute to The Turing Way? What would be the drivers for members of these groups? What are the barriers? Isla suggested that a barrier may be feeling un- or under-qualified and not wanting to bloat the material. She could imagine potential contributors being nervous of making unnecessary contributions and taking up reviewers' time or adding redundancy.

Konrad is wondering about business models that balance an open science approach and deliver a clinical tool. MELD has been built on open code and data from the start and the team feel strongly that this should remain the case…. but building a usable clinical tool (rather than “just” demonstrating the prototype principle) is a massive undertaking requiring many people and significant funding. Anyone know of any good examples?

Patricia is in the mood to run another ScienceTogether workshop. This was the event that she randomly attended in 2017 that got her into the SSI/RSE bubble: an unconference about collaborative tools. (The site isn’t online anymore, but you can checkout an archived version from https://web.archive.org/, or this blog post by David Perez-Suarez). The motivation is that its frustrating to see how few large projects have effective distributed collaboration practices. She’s thinking of running it entirely online (terribly meta 😆) and wonders what folks think about whether it would be useful and who would/should attend?

Georgia is trying to cut down her paper draft at the moment and is looking for pruning tips! 🌳️️️✂️🍃 She also asked about tips for attending MozFest for the first time. Patricia recommended making a schedule to try not to miss key sessions…but being ok with the fact that there will be conversations that you get pulled into that make staying really on top of things unlikely 😆 She and Kirstie both agreed that it’s pretty intense so be sure to take a time out whenever you need it. Also - the coffee and hot chocolate is free and delicious! 😋 ️🍫

Yini shared a recent paper, “Inception of memories that guide vocal learning in the songbird” (💰, no OA version), which describes how researchers have successfully implanted memories into birds, allowing them to learn new tunes without being taught. She’s wondering if this kind of technology could be possible for humans in the future? For example, for a person with post traumatic stress disorder (PTSD) patient the process could modify the patient’s memory in order to help them forget about the injury. Its an open question as to whether this approach would be ethical or not?

Ang wanted to know the lab’s views on a recent “matters arising” conversation about a paper on “Deep learning of aftershock patterns following large earthquakes” (💰, no OA version) which was published in August of this year. The comment (💰, no OA version) titled “One neuron versus deep learning in aftershock prediction” pointed out that a simple logistic classification regression did just as well as neural network models. The authors' replied (💰, no OA version) that these are philosophical differences on the goals of the modelling. Alternatively, a more cynical position may be that deep learning makes a paper more likely to be published in high impact journals than logistic regression 😒

Sarah is thinking of starting a blog and is looking for tips!

Elizabeth is wondering if anyone has suggestions for the ratio of papers-you-review to papers-you-publish? It can be a lot of work to review papers and she’s wondering what rubrics people use beyond “I think I could provide useful feedback to the authors.” Patricia shared a guideline of two papers for every paper you publish because your paper will need 2 reviewers (on average) and you should repay the favour to keep the system going. But the problem with that is that papers have multiple co-authors that could review too, so the math probably doesn’t quite work out. Kirstie recommends doing 4 per year and saying no to the rest unless you would want to lead a journal club on the paper (that’s the amount of work that goes into a review!)

World Mental Health Day

Although we didn’t have lab meeting in person today, the lab wanted to note World Mental Health Day. We collected together the following suggestions that we have each found valuable to support our mental health wellbeing.

  • Sharing examples of inclusive work practices can reduce stress - for example adding your work hours to your email signature.
  • Monitoring your hormone cycle can help with predicting and understanding pain or mood swings.
  • There are good online resources for self help: https://www.getselfhelp.co.uk and https://llttf.com.
  • Meditation has helped some of us in the past.
  • Seek medical advice as soon as possible and take medication when necessary. Anti-depressant, anti-anxiety and anti-psychotic drugs don’t work for everyone, but they can make a huge difference when they do.
  • Ask for information at your university or workplace. There are sometimes more resources - such as free councelling or health care support - than you imagine (and they can be hard to find on your own).
  • Remember that structural prejudice exists: it may not be you, it may be your environment making you feel bad.

The theme of this year’s World Mental Health Day was suicide prevention. There are some useful helplines and other resources listed on the NHS webpage. If you are concerned that you are at risk of suicide you should seek the advice and support of your GP. If you are in distress and need immediate help please call 999 or go to A&E and ask to speak to the nearest crisis resolution team.