22 October, 2020
Whenever I mention I organise hackathons, it becomes quite apparent that there is much confusion as to what a hackathon actually is, but it’s very simple: a group of coders get together for a few days to create new software or improve existing tools. Last year the Babraham Bioinformatics hackathon coders all met at the wonderfully retro Centre for Computing History in Cambridge, and we were all excited to return there this year. Except of course, we didn’t.
Like so many plans in 2020, we soon realised that an event involving scores of people working together indoors was a non-starter. I had resigned myself to the fact that there would be no hackathon this year, and I wasn’t all that keen on organising an online alternative. But after several months of barely leaving south Cambridge, the prospect of meeting other bioinformaticians, albeit via Zoom, became considerably more attractive. So, I set up a new website, an Eventbrite page and promoted the hackathon on Twitter and the usual Babraham Institute and Cambridge University channels. Within a few hours of launch we had the first sign-ups, and it was clear that an online event would be markedly different from the previous couple of codefests.
To my surprise, a week after launching, we had more bookings from Lagos in Nigeria than from Cambridge. Although the event was of course now open to everyone, I had marketed the hackathon as a Cambridge-area meeting, and yet people from outside the UK kept booking their places. In the end we had over 250 registrations, from UK, India, Nigeria, Uganda, Bangladesh, Ghana, Italy, Ukraine, Egypt, France, Hungary, Indonesia, Kenya, Mexico, Russia, Sweden, Turkey and South Africa. Most participants which were based outside the country, with India accounting for the largest contingent of people.
This year we had made the event free to all, and without the nominal attendance fee there could have been a large number of sign-ups who weren’t entirely committed to spending three days coding. I therefore was concerned on the morning of the first day that this promising but unexpected interest would not translate into actual participants. I needn’t have worried though, for, in terms of attendance, this was easily our most successful hackathon.
Many of the attendees worked on their own, but there were also eight structured project groups in which teams of people contributed towards a shared goal, interacting with one-another via Zoom and bespoke MS Team sub-groups. Perhaps the most prominent among these was the OpenVirus project in which a team of over a dozen people based in India built tools to mine the scientific literature for new insights into viral epidemics.
During the event most participants had free time to code and get on with their respective projects, although I organised breaks at various points in which people were randomly assigned to six-person Zoom virtual coffee tables. These proved an excellent way to meet others in a more relaxed setting. There was even a bioinformatics-themed pub quiz on one evening, which was more fun than it sounds!
During the hackathon the attendees created slides summarising their work and, as in previous years, at the end of the event they presented these to the rest of the group. Being on my own at home it was hard to gauge the level of work and exchange of ideas that was taking place in the event I was running, but the talks paid testament to the professionalism and commitment of the other participants. These presentations are now available on our YouTube channel and will be submitted to the online journal F1000 (projects from previous events are also available: 2018 Hackathon and the first Hackathon in 2017). The standout moment for me was a project that used algorithms to classify human lung X-ray results. It brought together researchers from South African and other countries, along with a Singaporean student who created a machine learning classifier. He was still at high school.
Obviously we are all desperately eager for life to return to normal, and I am looking forward to meeting people in person once again at a future hackathon, somewhere in Cambridge. But having said that, I will want to integrate virtual attendees into the event next time. This year brought home to me the countless individuals who are willing and certainly more than able to contribute to science, and with a laptop and an internet connection this is now possible.
Geographical separation from centres of excellence such as Cambridge is no longer a barrier to doing great collaborative work, which is certainly one positive note from an otherwise isolating year.
22 October 2020