At CMU Libraries’ biomedical hackathon with DNAnexus, open science and collaboration won

This past weekend, however, from Oct. 19 to 21, a hackathon hosted by Carnegie Mellon University Libraries and California-based cloud computing and bioinformatics company DNAnexus had the goal of collaboration and connection, instead of competition.

DNAnexus Principal Scientist Ben Busby, who is based in Pittsburgh, told Technical.ly that similar to past events hosted by the partners, the goal of this year’s hackathon was to bring the bioinformatics and biological data science community together.

“Everybody learns a lot, and then also, people network,” Busby said. “A lot of people get jobs through these types of events [because] in my opinion, it’s the best sort of technical interview anybody could ever have.”

Busby added that this style of hackathon was established by the National Institutes of Health in 2014. At the time, the NIH began running hackathons on basic subjects, but due to the rise of next-generation sequencing and genomics, he said, hackathons in the biomedical field became popular. This year, as a part of the theme of data management and graph extraction for large models in the biomedical space, participants were asked to work together on a different part of a problem.

This is the fourth hackathon of its kind that’s taken place at CMU, and Busby said the event averages about 40 participants each time. Depending on a team’s skill level, they could be asked to focus on assertion tracking in large models or flexible tokenization of complex data types. The end goal, Busby said, is for the hackathon participants to learn from each other.

Melanie Gainey, a CMU Libraries librarian for biological sciences, said since the participants are mostly — but not exclusively — academics, Ph.D. students, and graduate students, research is a natural part of the event. Throughout the hackathon, participants use publicly available datasets to create open-source pipelines and then get all of those outputs published on GitHub.

“One thing we realized very early on in the process, probably in our first or second hackathon, is that it is really critical to get librarians involved,” Busby said. “With almost all of our hackathons, whether they were at NIH academic institutions, or even working with Google or Amazon, we usually have some librarians involved.”

Gainey said that since librarians tend to have expertise in things like data management, documentation and publishing, the libraries were a natural partner to cohost the event. Additionally, part of Gainey’s role as the director of the libraries’ Open Science Program is to support open science and help researchers make their work reproducible and reusable, as well as foster interdisciplinary collaborations.

“This is a really great way to help researchers here actually practice open science and a really natural way for them to learn more,” Gainey said.

Through the years, the competition has been relatively successful in getting its results published on GitHub and even in scientific journals.

When the event was still being planned, Gainey and Busby recalled that there were a few hiccups — namely that several pros were supposed to come over from London to help put on the festivities, but ultimately couldn’t make it. Thankfully, Busby said, CMU and University of Pittsburgh students were kind enough to pick up the slack, which led to a successful event.

“The Pittsburgh community has stepped up,” Busby said. “I’ve been really pleased actually as a Pittsburgh resident that local folks were able to step up and really fill in some pretty big shoes left by some folks in London and elsewhere.”

This editorial article is a part of Biotech Month of Technical.ly’s editorial calendar.