Teaching genomics across Harvard schools
Awardees: Winston Hide (SPH), William Gelbart (FAS), Maxwell Heiman (HMS)
Summary: Awardees will establish the “Harvard Genomics Teaching Group,” share pedagogical approaches, a dedicated platform for analysis, and document best practices.
With the advent of inexpensive next-generation DNA sequencing, computer-based analysis of genomic data has become a core skill for molecular biologists. However, in-class analyses can be difficult to accomplish due to unpredictable run-time on over-taxed, shared servers, as well as scarce experimentation and research in effective pedagogy and assessment practices. Professors Max Heiman, Winston Hide, and William Gelbart set out to incorporate computational genomic approaches into what had been a traditional model organism-based genetics class. To address the associated challenges, they established a ready-to-use server running the free, open-source Galaxy genomic toolbox and created a “Harvard Genomics Teaching Group” to share pedagogical strategies for implementing these exercises in class.
The work, initiated in 2013 and ongoing today, was not without its challenges. According to Heiman, he and Curriculum Fellow Emily Gleason have experimented with different pedagogy and lesson organization over three semesters in Genetics 201 and are always evaluating and shaping the approach. In the first implementation, he found that it was challenging to include active elements into class time. “Just logistically—trying to run computer analysis on tiny desks, and paying attention to multiple things at once—it was hard for students,” he observed, and also heard from students in their course evaluations that the technology felt like a distraction. In 2014, he switched to a combined chalkboard-computer format where he gave a chalkboard lecture on underlying concepts, and interspersed the relevant analysis steps on a projector, which was better received by students and more engaging, but still a bit clumsy.
In 2015 he switched to a pure chalkboard lecture format, speaking to concepts and observations of the data analysis that students completed prior to class. In class, they made a table together showing how the different parameters they used affected the results of the analysis, showing that the global results would vary greatly but the strongest “hits” were robust to a wide range of parameters. “I feel like we finally got it right. I was very happy that, after quite a bit of trial and error, we seem to have found an approach to this material that works well for us,” Heiman said.
The following comments from student course evaluations reflect the progression of this workshop from a hands-on “practical applications” approach to a chalkboard “concept based” approach:
- 2013 (hands-on exercise in class, every student brings a laptop): “I think that it would have been a lot more helpful to go through first what we would be doing and why because just jumping right in we just had to follow Max and had no idea why we were doing what we were doing.”
- 2014 (instructor walks through example while students watch): “Did Max’s lecture really even count as a computation workshop? We were better off just reading that instruction sheet on the website than sit there in class.”
- 2015 (students do pre-class exercise on their own, instructor discusses concepts in class using chalkboard lecture): “The written guide was particularly good and I liked how class time was spent on conceptual explanation of the techniques (rather than walking through the guide’s prompts to go to the website and do this, etc).”