From high school to postdoc: Lessons from a decade of bioinformatics education

As a postdoctoral research fellow with both a PhD and a bachelor’s degree in bioinformatics, my scientific background is the product of over a decade of bioinformatics training and education. Given the advent of more efficient and higher-throughput experimental approaches and the ever-decreasing cost of generating and storing biological data, there is a greater need than ever before for highly-trained bioinformaticians, both in academia and in industry. A high-quality, successful bioinformatics educational experience needs to address two key components of scientific research. Bioinformatics education needs to first and foremost cultivate a deep understanding of and appreciation for basic biological mechanisms and concepts. This understanding should include not only the current state of biological knowledge, but also the historical account of how that knowledge was obtained. Without establishing biology-centric motivations for asking “why”, bioinformatics trainees are in danger of becoming black boxes for performing analyses without a connection to the interpretation of results. The second requirement for bioinformatics education is promoting the ability to think computationally and understand complex mathematical and statistical concepts as they relate to biological data. It is no longer sufficient to be able to run computational programs in the command line or write simple python scripts to perform a computational analysis. Like the biological aspect, the mathematical and computational aspect of bioinformatics training needs to address questions of “why”. For example, why does a certain distribution model better fit a particular type of biological data? Too often bioinformatics education teaches the “how” of computational approaches but not the “why”, which again can lead to a black box of analyses that actually do not make mathematical sense. Many students entering graduate bioinformatics programs have been trained in only one of the two key areas and lack the required depth of knowledge in the other field to be successful bioinformaticians. Graduate bioinformatics programs often require new students to train intensely in the lacking field for the first few semesters, which may not ensure that an adequate understanding in both fields is achieved. In this talk, I will argue for bioinformatics education to start at the undergraduate and even the high school level, starting with teaching the fundamentals of biological, mathematical, and computational theories. Upon these theoretical foundations, then, can be built the knowledge of how to design and perform experiments that address questions being asked. I will also outline specific examples from my perspective of bioinformatics training that is being done well and what still needs improvement.