PhDs and Pattern Formation: My Summer Research Project

Isn’t it incredible how out of the randomness of life, we find patterns all over nature? This summer…

Ok I’ll stop - I’ve said that sentence enough times this summer already!

For those who don’t know, this summer I was selected for the Mary Lister McCammon Fellowship, which is a 10 week funded research programme with Imperial College London for females thinking about applying for a PhD this year. Thirteen students from many different UK universities were selected and assigned a supervisor who aligned with their research interests. At the undergraduate level, the male to female split of students is fairly even. However, by the time you get to PhD level, the percentage of females drops to around 20%. This is the reason why a number of the PhD programmes associated with Imperial agreed to fund such a programme, to help and inspire more women to pursue research beyond the Master’s level.

I had been pretty vague on my application about my research interests and I decided not to select any preferred supervisors, as I was really unsure about what I wanted to do. So I thought I would let them decide for me. After a two month wait, I found out I had been selected for the programme and I had been matched with Dr Thibault Bertrand. His research on non-equilibrium statistical mechanics (don’t worry I’ll explain all these long words in a minute) actually seemed really interesting. I hadn’t really considered him on the list of supervisors before this, so I’m glad that I let fate decide for me!

Soon afterwards, Thibault and I exchanged a few emails and agreed to have a preliminary meeting to discuss the project in May. He also brought along Dr Gunnar Pruessner as a second supervisor. They both seemed nice and I was excited about the project they were proposing on Phase Separation in Biological Tissues, even if I still didn’t really know what all these words meant.

Two years ago, when this programme started, the whole project took place in London. However, for obvious reasons, last year it was fully online, and it was looking like this year’s would be too. I decided to start the project remotely (especially since all our cohort activities would take place online), with the idea of perhaps travelling down to London for a week or two towards the end of the summer. I also hadn’t decided how much time to spend in my Edinburgh flat over the summer and at home in Linlithgow. In the end, I was in Linlithgow with my family for most of the summer and went into Edinburgh every couple of weekends to see friends.

My project

For a gentle introduction to my project, watch this video.

Essentially I was using particle/ agent based modelling to model pattern formation in pigment cells. It's in the area of biophysics, and in particular 'non-equilibrium statistic mechanics', as I was dealing with so-called 'active matter' systems, in which energy is constantly being used up. This leads to the system being out of (thermodynamic) equilibrium. These systems display interesting phenomena that you don't always see in equilibrium systems. The statistical mechanics part just means applying statistics/ probability based methods to describe physical systems macroscopically.

In the end, most of my project actually related more to the physics side of things than the biology side of things, but I still enjoyed having the biological questions for background motivation.

What this translated to day-to-day was mostly a lot of coding (in Python). I spent a few weeks reading up on the area, and replicating results from the literature. I then made a new model with two populations of particles.

Speed was a big constraint so I spent a long time trying to optimise my simulations using the Python package Numba, optimised neighbour list searching methods, trying to learn C before deciding that Python with Numba was going to be easier and nearly as fast, and using the Maths department's computer cluster (so I could leave many scripts at once to run on their servers and not on my poor computer).

Once I had all that working, I would set up my models and scripts, leave them to run on the cluster, then analyse them, before sending more scripts to the queue on the cluster.

By the end, I did have some interesting results that don't seem to have been discussed in the literature before, and could be written about in a short paper if I had more time. This left many open questions, but I think I did pretty well to get that far in 10 weeks.

You've done enough reading now to earn a cool animation:

If you're still interested in learning more, you can read all the gory details here.

Week 1

By the end of June, I’d had over a month off from maths and uni work. I therefore found the first week of the project very tiring as I wasn’t used to sitting down for 6-7 hours of working. On top of that, this was a scary new project with lots of information to take in, which was rather overwhelming. In order to keep myself motivated during this first week and to remember what I’d actually done each day, I decided to start a little ‘research log’ in a Microsoft Word file. With research, progress can sometimes feel slow. Some days, you make lots of progress and other days you don’t understand anything. This research log helped to show me how far I had actually come and that I did make a lot of progress, even if it didn’t always feel like that day to day. By the end of the first week, motivation had quickly grown once I got started with some Python coding. I really enjoyed those first few weeks of the project.

Supervisors

I met with my supervisors twice a week for one hour. I was very grateful to be given this much time with them. Usually PhD students and supervisors wouldn’t meet this often, but Gunnar suggested this, as 10 weeks isn’t a very long time in the research world, so this would help me not to stay on track and not get too lost. In any case, I was very grateful to be given this much individual time with them! From very early on, I really got on with Thibault. I always came out of our supervision meeting feeling inspired, even if I was feeling rather stuck and demotivated just beforehand. Even though Gunnar had suggested the twice weekly meetings, he didn’t actually turn up to half of the meetings, and then he went on holiday for the whole of August…! Anyway, even though I didn’t spend as much time with him, he helpfully met up with my during the week that Thibault was away on holiday for a week. I also warmed to him as the project went on.

Every week, Thibault also hosted a group meeting, with all his students (1 post-doc, 2 PhD students, 2 Msc students, an undergrad group summer project with 3 students, me, plus a few others who were currently working or had worked with Thibault who still came to group meetings!). He would go round the group to see how we were all doing. Then, each week, we would take it in turns to present something about our most recent results or present a paper from the literature. It was nice to be able to interact (online) with other members of his group, although the group meetings often went on for 2+ hours. We also had a Slack group where I could message him or his other students for help.

Nearer the end of my summer it was my turn to present to the group. I think this was really good practice and it was nice to be able to prepare a more maths-y talk about my project. I did a similar presentation later in September for Gunnar's group too, when I had more results to talk about.

Cohort

As a cohort of students part of the Mary Lister programme, we had a 2 hour communications training every week with Justine Jones. These led up to a final 5 minute, 5 slide presentation for a general audience (vaguely inspired by the three minute thesis talks). However, for the first half of the programme, these trainings were ‘applied improvisation’ classes – essentially just a drama class on Zoom to break the ice and get us comfortable with talking. I think these would have worked much better in-person, as it all felt a bit strange on Zoom. But it did make me think how you can use your voice in different ways to convey different feelings, even on a computer. We also discussed the importance of communicating science and your research to the general public.

For the second half of the project, each week we would practice parts of each of our presentations every week and give lots of feedback to each other. Many of the cohort members were doing very pure/ abstract topics, with very little connection to the real world, which made this lay presentation a lot harder for them. This provoked many debates and interesting discussions about how to communicate a complicated topic without making the talk feel too simple or silly. The main takeaways for me were to try to get people excited about your project when they listen to your talk, and to give a general gist of your project without too many complicated equations and details.

On the last Friday afternoon of the project we presented these to all of the supervisors on Zoom. We had to come up with catchy titles and mine was "How Did the Zebrafish Get Its Stripes?". The final talk is the one I linked above for a gentle introduction to my project. Here’s the link again if you haven’t watched it yet.

We also had a weekly coffee meeting with the programme coordinator, Dr David Ham, to chat informally about everything Imperial and PhD related. Occasionally, we also had seminars/ talks by current female PhD students at Imperial. Being on Zoom, there were sometime some awkward silences. But some of us started to stay on the call after David left, which was when we really got to chat and get to know each other. We also had a WhatsApp chat for communicating in between coffee meetings. It was refreshing to meet so many like-minded women, all very motivated by mathematics research.

Digital nomad

Due to Covid restrictions, the programme was mostly done remotely. In the end, it was actually a nice thing to spend more time at home (and not pay 10 weeks worth of London rent!). Another perk was being able to go on holiday with my mum, brother and his two friends (+ dog) to Mallaig on the Northwest coast of Scotland for a week in a holiday cottage to go to the Scottish 6 Days of orienteering! Luckily the broadband was working (as I had no signal on my phone) and we had a lovely cottage which as great to work from all week. I took one day off and so managed to fit in three days of orienteering in between my project work. A highlight was orienteering on the Isle of Skye. It was nice to be able to pack all my work into a small rucksack and have a change of scene.

London!

I was still hesitant about going to London until the week before I left. I’d waited until I was double vaccinated so felt a bit safer than before. Fortunately for me, my first cousin once removed lives in a lovely house in Chiswick in London with her family (+dog), which is ideally located for Imperial College London’s South Kensington campus. Only 12 minutes on the District Line (one of the nicer tube lines) and about 10 mins walking either side. I stayed in their house for two weeks. For part of the second week they were actually on holiday so I had a 4 bedroom house in Chiswick to myself for nearly a week – no complaints!

On the very first Monday I was in London, working from my bedroom on my computer, I was downloading my data files after my Python scripts had been running over the weekend. I had run some larger simulations after implementing some faster methods and thought I was just going to the same thing as I had been seeing for several weeks – no patterns or clusters. But at 9:30am on the first day working in London I finally found what I had been looking for: phase separation! I was so excited and didn’t know who to tell so I went downstairs to find 9 year old Charlie, who was watching TV. I showed him my particles snapshot telling him how excited I was, but he wasn’t that bothered.

During my first week there, I worked a few half days in the Maths building’s Maths Learning Centre, which was pretty empty over the summer and so our cohort had adopted it as our main study space on campus. It was lovely to finally meet some of the cohort and each lunch together and talk. One evening it was one of their birthdays so we had a spontaneous picnic in Hyde park after work in the dark. Very wholesome. That evening there were a record 9 of us (7 cohort + 2 others) all together in person. On the last day I did my presentations from the MLC with a few others so we could celebrate together afterwards at the Student Union bar and finally meet the programme coordinator.

Being in London meant I was also able to meet my main supervisor, Thibault, in person! In the first week we met up in the Maths building so he could show me round and go out for coffee. I was expecting this to take no more than an hour but we were out for 2 ½ hours in the end! We walked around campus with one of his PhD students, Henry, before going to the V&A museum café where he bought us drinks. In the second week, we met up with a few more of his students at a café/ bar across the road from Imperial for another 2 ½ chat after work! I’m very grateful for all the time he spent with me and showing me campus – all our chats were really helpful.

Although Imperial’s campus is in an amazing area of London (South Kensington), the actual campus itself is not that nice. However, Thibault brought me to a very nice balcony terrace on the 8th floor of the Physics building with views all over London. You can also see the Royal Albert Hall from there really well. I think it’s the best place on campus. Somehow, I managed to find my way through the maze of the joint Maths/ Computing/ Physics building back there again with a few of my cohort twice more to eat lunch.

Apart from working on my project, it was nice to have a small London experience at the end of my project to get a proper taste of the city and student experience there. I was pretty convinced I would not consider it as a place to live before coming, but now I’m not so sure. I am now considering it as an option, especially since I got on so well with my supervisors, but I’m still a bit hesitant about living there. In any case, I’m glad I did manage to go. I got to know the local area quite well, but I also managed to: visit friends in Primrose hill, cycle around, visit the V&A and Science Museum, visit Covent Gardens, go to Kew Gardens. My cousin Juliette also happened to be in London during my last weekend and it was nice to spend the day with her after not seeing each other for 10 months!

Oxford & Bath

When planning this trip, I suddenly realised it would be good to take the opportunity to visit other potential PhD cities while I was down south. So on the way down I spent a weekend in Oxford, and then after my London trip I spent a couple of nights in Bath. I had a nice time at both of these places, and I was happy to have a little bit of solo travel and excitement to finish off my summer (even if it was just a domestic Interrail trip this time).

In Oxford, I stayed in one of the colleges for the authentic experience (plus all the colleges were closed to visitors due to Covid unless you were staying in their accommodation). I mostly walked around a lot. I was reading the book Magpie Lane at the time, which is set in Oxford, so I also liked seeing all the places mentioned in the book in real life. I also attempted to use the the bike hire scheme (not a great experience as it was too expensive and my bike had a flat tyre).

Bath was also nice to have a bit of a detox after all the hustle and bustle of London. I was very fortunate with the weather there; sunny and 25-28C during the day! I've never experienced September like it as I've always been in Scotland that time of year. The uni was a bit of a trek up a very steep hill (not really cycle-able), but the city felt like a (steeper) mini Edinburgh in some ways. It has to be said I did visit Bath without actually visiting the Roman Baths, so I'll have to return (or do a PhD there!) to go to them. I did however go on a free walking tour of the city to learn more about the history.

What next?

This summer programme really confirmed to me that I would like to pursue a PhD next year, so my next semester is going to be busy with applying to places! I still need to make my mind up about all the different options as there are so many: research areas, supervisors, projects, cities.

Thank you for reading and don't forget to subscribe to my blog down at the bottom of the page to get any updates directly to your inbox!