Merging Biochemical and Analytical Training: a Computational Approach to Chemical Communication

Abstract

As techniques in biochemistry and neuroscience progress, so is the volume of data generated by advanced technologies. Therefore, there is a need for training young scientists in both fundamental biology and biochemical concepts, but also in the handling, processing, and analysis of large datasets, i.e., Big Data. Two years ago, in the Department of Biochemistry at Virginia Tech, we created a new course entitled “Biochemical Communication,” where senior undergraduate students learn about the biochemical and evolutionary principles guiding chemical communication between organisms. Core components of the course are: 1) Lectures relying on active learning strategies, 2) student-led weekly discussions of the primary literature, and 3) a semester-long project consisting of analyzing large data sets generated in the laboratory of the instructor. Over the past two years, the data sets provided were: electrophysiological recordings from neurons of mosquito antennal lobes (i.e., primary olfactory centers) and gas-chromatography mass-spectrometry data resulting from the collection of plant headspace volatiles. Over the course of the semester (15 weeks), students acquire a working knowledge of data manipulation and visualization techniques, descriptive and inferential statistics, multivariate analysis, and machine learning approaches. These methods are deployed with the programming language R. Students’ training, as well as the approach they elect to employ, are guided by their drive to answer a biological question of their choosing. The work is conducted in groups to promote peer-mentoring. At the end of the semester, each group submits a seven-page long report structured as a scientific publication. The feedback collected after completion of the course indicated that students are receptive and enthusiastic about such hybrid classes where an introduction to biological and biochemical concepts is merged with hands-on learning of programming and data analysis.