This post contributed by Jake Lever, a PhD student in the University of British Columbia’s Bioinformatics program. NumFOCUS was pleased to provide funding support to the hackseq17 hackathon.

hackseq17: Canada’s genomics hackathon

hackseq17: Canada’s genomics hackathon was held at the University of British Columbia (UBC) in late October. This event brought together a diverse set of researchers to solve pressing problems in bioinformatics and genomics. The goals of hackseq include encouraging open science, improving diversity, starting collaborations between researchers and educating researchers about new areas of genomics.

Last year’s event was a resounding success and had kicked off some exciting collaborations between researchers that continue to this day. With the support of NumFOCUS, we wanted to start the hackseq tradition and bring another group of enthusiastic researchers together.

This year’s event took place over three days in the Life Sciences Centre at UBC. Many participants were undergraduate and graduate students from the two largest universities in Vancouver: UBC and Simon Fraser. We also had post-doctoral researchers, researchers from industries and staff scientists in academia from British Columbia and the Pacific Northwest.

We were particularly excited this year to work with virtual participants. These are researchers from around the world that would work remotely on a project and communicate with their team members through Slack, a popular messaging app for collaboration. Even with some drastic time differences, we had great researchers working around the clock from France, the UK, India, China and many other countries.

Months before the event, we had put out a “call for projects” to research groups across the Pacific Northwest. Eight projects were selected, led by post-doctoral researchers and senior graduate students. The projects were as diverse as the participants:

• The Biosyntax project looked at syntax highlighting for bioinformatics data
• The SCREW project aimed to build a reproducible workflow for single-cell epigenetics data
• The Fusebench project focussed on the problem of understanding gene fusions, when different tools give provide results in different file formats
• The ShinyDiversity project build an R Shiny interactive app for examining diversity statistics from ecological data
• One project focussed on the steep learning curve for the R tidyverse set of data analysis and visualization tools and worked to build better tutorials
• The ChromeQC project focussed on quality control for the new Linked Read technology from 10x Genomics
• One project built a machine learning pipeline for predicting efficiency of sgRNA sequences in CRISPR technology
• The CollapseTree project aimed to improve phylogenetic tree visualization by building tools to intelligently collapse portions of a tree

On the first day, the teams came in and met each other. Some groups had already been communicating through Github for weeks but had not actually met yet in person. The virtual participants also joined in remotely by Slack. Some groups also elected to use video conferencing for the planning sessions. Teams were encouraged to focus on planning for the first morning so that all participants understood the scientific question and their role in the solution. Then in the afternoon, the hacking really kicked off.

The venue at the Life Sciences Centre offered an incredible open atrium space which provided the perfect space for collaboration. Plus the table of coffee, snacks and lunch in the middle kept all teams sustained through the whole event.

The second day of a hackathon tends to be the most relaxed. Teams are hitting their groove and the end is still a whole day away. Many teams had working prototypes towards the end of the second day. In the evening, we held our official hackseq social at a nearby pub. This was a great opportunity for teams to get to know each other better and expand their bioinformatics network.

The third day is really crunch-time. The teams are putting their final touches to the projects and making some nice figures to show off to all the groups. At the end of the day, all the teams give short 5-minute presentations describing the challenge and their work. The talks were webcast so that virtual participants and other interested persons could watch the talks. We also had a community vote for the best teams. Teams Biosyntax and Fusebench were chosen as the first and second place winners for the People’s Choice award. Congratulations to them!

Throughout the hackathon, we encourage the teams to work on short papers to explain their projects. Through sponsorship, we offer to pay the publication fees for a short paper on the open-access F1000 hackathons channel. Each team nominated one member to be the official writer and led the writing of their short paper. We’re happy to say that two teams have almost complete papers and are beginning the submission process. Several other teams are close to a finished paper as well. Plus all the code is up on Github!

So thank you to NumFOCUS, as well as our other sponsors: ECOSCOPE, GenomeBC, MSFHR, ACM-W, RStudio, New England Biolabs, UBC AMS, Chipotle and Bioinformatics.ca.

We look forward to hackseq 2018!

—

Help us support more events like hackseq — become a NumFOCUS Member!