This summer, we were delighted to welcome seven interns to the Unit – Laban Masunga, Benjamin Pitter-Fagan, Davina McLaverty, Weishi Chen, Luka Kovacevic, Ami Saito and Anqi Sui.
Our internships provide those with an interest in data science and knowledge of statistics to join the Unit to work on projects being led by BSU researchers. They carry out data analysis, take part in research team meetings, present their findings, and get involved in all aspects of life at the BSU.
For Laban, Benjamin and Davina, their internships were part of the HDR-UK Black Internship Programme – giving paid work experience to future Black data scientists and bringing new perspectives to the work already being done in data science. Benjamin, Laban and Davina were each supervised by a different member of the Unit and worked on projects ranging from statistical genomics to Bayesian modelling.
For Weishi, Luka, Ami, and Anqi, their internships were specifically targeted to working in the DART research theme at the BSU, getting involved in a real-life problem of designing a Phase I dose-escalation study motivated by an ongoing trial on COVID-19 treatments. Their internships were administered by the University of Cambridge’s temporary employment service.
This year, all our recruits carried out their internships remotely due to COVID-19 guidelines. This provided new challenges for both our interns and our staff, however we worked to ensure they were fully integrated into the Unit and were set-up with all the practical elements needed.
So what did some of our summer 2021 interns have to say about their experiences?
“My internship, under the coordination of HDR-UK, was a fascinating opportunity and life-changing experience.
Before joining the BSU, I was nervous and thought things would be difficult because of my limited knowledge in health data science and specifically Genomics, where my project was centred. The project involved modelling relapse of breast cancer using a high-dimensional genomics dataset (a dataset with more than 35,000 variables) from more than 2,000 breast cancer patients for the METABRIC cohort, a survival analysis case.
Special thanks to Oscar Rueda for his trust in me. I remember I met him for the first time during the interview process and after joining the BSU he made my journey so easy and cherishable, taking me step by step and ensuring I’m getting things right. Also, much thanks to Solon Karapanagiotis for his constructive ideas in the last few weeks indeed was a great honour having him aboard.
My R programming has significantly improved during my internship, and I am much more confident in playing with any big dataset no matter how unarranged it could be; indeed, it was enjoyable to access and see how high performing clustering (HPC) machines work to simplify my computational tasks during multiple sampling stage in my project.
Enormous applause to the HDR-UK for coming up with this initiative of Black interns which has completely changed my understanding of how diverse Data Science is especially by introducing me to Health Data Science, an area that I’m starting to grow a high interest in. Undoubtedly, it was a great honour to be at the BSU through HDR-UK, and I will remember and cherish this internship in every aspect of my profession.”
“Embarking on my health data analyst internship with the BSU was a great experience. It was a brilliant opportunity, as it was my first major role in the data analysis industry. I was introduced to everyone in the team, with members explaining their day-to-day roles to me. The IT team helped me with setting up all of my login details, establishing me on their system remotely and were very patient with me throughout time, resolving any technical difficulties. Being able to work at such a prestigious institution has given me a major career boost going forward.
I was working under Jennifer Asimit on a multi-trait fine mapping project, which involved analysing 4 lipid traits for chromosomes of European ancestry. Jennifer was a very supportive supervisor and assisted me with everything I needed to succeed during my time. Parallel analysis was being run on samples of African ancestry, by scientists whom we were collaborating with in South Africa. My task was to find potential causal variants for genetic associations, which were then fine-mapped, to be later merged with the results from the African samples, via a process called trans-ancestry flashfm. I learnt R programming as a beginner, before starting this internship and really grew into it, as I progressed and had to use R to handle really large datasets. I was invited to weekly BSU meetings and really felt like a part of the team.
HDR-UK had an excellent programme overall, as we had weekly career talks and institutions such as the NHS and British Heart Foundation talking to us about their work and career prospects; with a great platform to network.”
“Biostatistics is an area that I knew little about before starting my internship. I always knew that I loved statistics but only had very vague ideas about the area I wanted to specialise in. Hence, when I came upon this internship, I thought this was a great opportunity for me to explore a new area.
The first three weeks give an introduction to the phase I clinical trial. I read about classic designs and some further adjustments of dose-toxicity trails, in particular, the CRM (continual reassessment method) and partial order CRM (POCRM) designs to determine the optimal dose regimens. Then, I was introduced to the AGILE platform, which is a multi-central seamless phase I/II platform for COVID-19 treatments. Motivated by features of the AGILE platform, my task is to develop a partial order CRM for a two-parameter logistic model.
More importantly, in addition to the knowledge side, this internship helped me to discover the area of biostatistics. I got to know the life of a biostatistician, which involves lots of reading, coding, collaborating with pharmaceutical companies, etc. It also changed my view that biostatisticians only do very applied works. In fact, there are lots of theoretical and methodological works in biostatistics.
Despite being in the middle of the pandemic and having to work remotely, I feel the connection with people at the BSU was brilliant. The interns were given career talks from people working at the Unit, which included professors, researchers, and PhD students. We also had the chance to attend the BSU Together meetings. In addition, there are also weekly social calls where people can chat freely.
To summarise, I had the most enjoyable experience working as an intern at the BSU, and my internship could potentially be life-changing! I will definitely apply for a PhD at the Unit, so hopefully, this is not yet the time to say goodbye.”
“Prior to starting my internship at the BSU, I was always interested in the ongoing research at the Unit and would regularly check their pages for opportunities. I was thrilled to find this opportunity advertised on Twitter, and following the interview along with what felt like the longest four days of my life, I was honoured to be offered a place!
The first few weeks were spent studying Bayesian approaches to Phase I clinical trials and optimal dose-finding along with some of my fellow interns, where we did a literature review and replicated existing simulation studies. After learning the basics of this fascinating subfield, we were each assigned our own projects where I focused on implementing Bayesian model averaging to take into account further uncertainties involved in dose-finding.
The research environment at the BSU is one that, despite being completely virtual due to COVID-19, I felt was especially warm and inviting as we had numerous talks from researchers at the Unit where we learned about their research and unique paths to the BSU. My supervisors, Helen Barnett and Pavel Mozgunov, were both incredibly supportive and always responded to my emails with swift and detailed replies to my questions. They also gave me feedback on my work which taught me how to justify novel methodologies and communicate them effectively to ensure that they are convincing enough to encourage people to use them.
This internship is one that I believe has helped me mature as a researcher and I genuinely hope that I will have the opportunity to be a part of this wonderful community once again!”
“Having just graduated from UCL with a bachelors in statistics, this internship was the perfect bridge to my master’s in health data science at LSHTM. My project with Haiyan Zheng was to develop a novel methodology to incorporate both efficacy and toxicity data to decide on whether to continue or terminate early-stage clinical trials. We used copula functions to construct the joint probability of efficacy and toxicity and then used Bayesian inference to update our prior beliefs on these probabilities. I particularly enjoyed learning OpenBUGS to run MCMC sampling methods and using the R2OpenBUGS package to call these functions from R. My weekly meetings with Haiyan were something to look forward to every week, as she was always encouraging and supportive throughout the entire journey of this internship.
In addition to the project, the weekly events hosted by the NIHR was a great place to learn and socialise with interns working at different universities. By inviting professionals in the fields, the events featured multiple topics such as meta-analysis, research methods, career prospects, and fellowships offered by the NIHR.
Throughout this internship, I have gained valuable experience, networks, and a glimpse into how it might be to do a PhD. I would like to thank everyone who supported me during this summer, and I wish the best to the future interns!”