Please join the eScience Institute on Wednesday, February 5th at 4:30 pm in the Gates Commons (CSE 691). Refreshments will be provided. PLEASE NOTE NEW SEMINAR LOCATION.
Joseph L. Hellerstein (Google):
Lessons Learned from Teaching Biochemistry to Computer Scientists
In the Fall of 2013, I taught an introductory course on biochemistry as a CSE 599 class. My objective was to help CS students and faculty acquire knowledge that will allow them to operate as peers with researchers in the life sciences. The course focused on proteins, enzymes, and a molecular understanding of the “central dogma” of biology – replication, transcription, and translation.
At first glance, biochemistry seems far afield from computer science, especially for students with no recent background in chemistry (and who have never taken organic chemistry, the standard pre-requisite for biochemistry). However, I discovered that much of biochemical knowledge involves concepts and tools that are an integral part of computer science.
First, a lot of the difficulty with learning biochemistry is that it deals with complex structures such as large molecules, cells, and organs. It turns out that tools such as the Unified Modeling Language (UML) are extremely effective for representing complex biochemical structures. Indeed, my class often had great discussions about the design of UML structures for proteins and metabolites; these discussions were an excellent vehicle for understanding concepts in biochemistry.
A second impediment to learning biochemistry, in my opinion, is that it is taught with an impoverished discussion of the computational problems being addressed by life processes. One example of this is cellular pathways. These are workflows. But I have never seen biochemists use the rich representations and tools of computer science in their discussion of pathways. Another example of computational problems solved by life processes is cellular transport of molecules, such as the transport of secretory proteins. Much of the complexity here is about the way addressing and routing works, something that is fairly digestible to a mature computer science student.
This lecture will discuss the pedagogy that worked well, what didn’t work, and how I plan to radically restructure the class when I teach it in Spring quarter.
Bio
Joseph Hellerstein is an Engineering Manager at Google, Seattle and a Senior Data Science Fellow with eSciences at the University of Washington.
Upcoming Seminars:
* February 12, 4:30 PM (CSE 691)
Joel Zylberberg (UW Applied Mathematics)
Computational Neuroscience: from the top-down, the bottom-up and everything in between
* February 19, 4:30 PM (CSE 691)
Steven Roberts (UW Aquatic & Fishery Sciences)
TBD
* March 5, 4:30 PM (CSE 691)
Chris Bretherton (UW Atmospheric Science and Applied Mathematics)
Big Data meets Big Models: Weather Forecasting and Climate Modeling