The title of this blog is derived, via a circuitous route, from a famous quote in the world of modeling (don’t worry if you’ve never heard it before — fame is relative) attributed to John von Neumann by Enrico Fermi:
“In desperation I asked Fermi whether he was not impressed by the agreement between our calculated numbers and his measured numbers. He replied, “How many arbitrary parameters did you use for your calculations?” I thought for a moment about our cut-off procedures and said, “Four.” He said, “I remember my friend Johnny von Neumann used to say, with four parameters I can fit an elephant, and with five I can make him wiggle his trunk.” With that, the conversation was over.'” Freeman Dyson, Nature 427, 297; 2004.
But apparently it takes more than 4 parameters. The number is actually more like 30 just to draw the elephant — according to Wel (1975) via a post by mahalanobis. The question of how many parameters it takes for trunk-wiggling is left as an exercise for the reader.
The moral(s) of the story: biology is more complicated than you think. And you can’t tell how many different factors are going to be important until you do the experiment.
This blog is about the interests of the Department of Systems Biology at Harvard Medical School. We use tools from physics, mathematics and computer science to help us better understand the behavior of biological systems, large and small. Our interests therefore include:
Methods for quantitative measurement, and for data analysis. Although much is said about the flood of new data in biology, nearly every time you want to understand a biological system at a mathematical or mechanical level you find that the numbers you need most are missing. Measuring and extracting the parameters that describe key features of the system is a major interest.
Theoretical and computational methods that can cope with the special features of biological systems. Issues such as combinatorial complexity, stochasticity, and variation from individual to individual and tissue to tissue are hard to deal with using conventional tools.
Philosophies of modeling. How do we represent what we know about the system — what level of abstraction is appropriate for a given question, what is important and what can be ignored? What are models useful for?
Evolution. One of the more useful tools to identify what’s important is evolution — a comparison across species helps to show what is “allowed” to change and what is not.
Synthetic biology. If I understand it, can I build it? (And in any case, can I build useful stuff?)
We will be posting thoughts about recent papers in the literature that we find interesting, news about the Department, and information about Department Alumni. Please check back frequently, and feel free to comment.
— Becky Ward
PS: See this post for what the blog turned out to be about in its first 7 months of operation.