As a mathematics professor at a ‘primarily undergraduate institution’ (a.k.a. PUI) who has been deeply involved in some local and national efforts at reforming undergraduate biology and mathematics education to be more interconnected, I delight at coming across papers with titles like this one: “Computing Has Changed Biology - Biology Education Must Catch Up” by Pavel Pevzner and Ron Shamir (Science 31 July 2009: Vol. 325. no. 5940, pp. 541 - 542). Often times, papers like these are written by people who are passionate about STEM education reform, and they share in their writings insights and perspectives that are useful to me.

This is not really one of those papers. In this article, the authors make an ineffectual argument for changing the undergraduate biology curriculum to include more training in computational biology, a.k.a. bioinformatics. It’s not that what they want isn’t reasonable or that it’s not needed. The authors do want to reform the undergraduate biology curriculum in a meaningful way. It’s just that this note in the Education Forum of Science is more of an excuse to report on a conference: RECOMB Bioinformatics Education Conference (http://casb.ucsd.edu/bioed/). Mostly, the authors repeat many of lines we’ve heard from many others since the publication of Bio2010.

This article does give some decent concrete ideas for how a computational biology course would be designed. For example, a course could introduce the computational content/concepts in the context of framing questions such as “Did our ancestors interbreed with Neanderthals?” or “How do we distinguish between different forms of breast cancer and choose the appropriate chemotherapy?” or “How can biomarkers be used to predict clinical outcomes of young breast cancer patients?” But the authors don’t go far beyond that. And this kind of suggestion, of teaching new concepts in a context that is immediately meaningful to students, is not innovative or new; educators that specialize in how students learn and in broadening participation in science and mathematics (cf. the BioQuest Curriculum Consortium at http://bioquest.org) have been recommending this approach for many years,

That the authors are ignorant of the work others have done in student learning and pedagogy reform is understandable (though not necessarily excusable, given the aim of this article). That I don’t understand is the authors’ apparent ignorance of or indifference to the challenge of adding anything to the standard undergraduate biology curriculum. Degree programs in the life sciences are already tight, and requirements set by admissions to medical schools further limit program flexibility (like it or not). Perhaps the recent AAMC/HHMI Committee report “Scientific Foundations for Future Physicians” (see http://www.hhmi.org/news/SFFP20090604.html) will lead medical school admissions committees to raise their expectations of applicant competence in mathematics, statistics, and computation. Such a move would open the possibility of biology programs requiring the type of new course proposed by Pevzner and Shamir. But even if medical schools calls for more quantitative sophistication (or literacy) from medical school candidates, we should ask if a course in bioinformatics should be required of all biology majors. That’s a discussion I’ll leave for others and for later.

In the end, the authors say that implementing their so-called proposed course “may become a first step toward building the new computational curriculum for biologists” without ever proposing a concrete course. At best, they give the readers the “pedagogical challenge” of designing this course so that it “(i) assumes few computational prerequisites, (ii) assumes no knowledge of programming, and (iii) instills in students a meaningful understanding of computational ideas and ensures that they are able to apply them.” At worst, they come across as people who have little knowledge of or appreciation for the tightness of the undergraduate biology curriculum They can parrot the spirit of Bio2010 well, but they’re not offering much meat for those of us who are in the trenches trying to make things happen.

Before closing, I’m going to pick a nit with their imperial use of the adjective ‘computational’ when used to modify ‘biology.’ While many people have used the terms ‘bioinformatics’ and ‘computational biology’ interchangeably. This was reasonable at a time when genomics led the way in leveraging computational and algorithmic power to illuminate questions in biology. But now applications of computation can be found in every corner of in the life sciences. Agent-based models illuminate ecosystem dynamics. CompuCell’s pde-based methods are used to model cellular phenomena. Computer-assisted tools use mathematical models of shape and image-analytic techniques to understand medical images. These are just a couple examples that have come up in Truman’s undergraduate mathematical biology program. Each of these are examples of ‘computational biology’, where computational is used in its broadest sense to mean an application of computers and computational methods. Let’s relegate the synonymous use of ‘bioinformatics’ and ‘computational biology’ to the history books with a nod and a thanks.

With all that said, and as much as the article didn’t thrill me like Joel Cohen’s “Mathematics Is Biology’s Next Microscope, Only Better; Biology Is Mathematics’ Next Physics, Only Better” (see http://www.plosbiology.org/article/info:doi/10.1371/journal.pbio.0020439), the short article is worth reading, if only to see that there are many people out there who want undergraduate curricular reform. And the references are decent.