Humming away in the Rutherford Physics building, a long cold walk from Stewart Bio, is a computer that can predict one of the fundamental processes in biology: how vertebrae form.
Paul François, associate professor in the department of physics, and associate member of the department of biology, is one researcher applying physics to biology problems.
“There is a current boom in this field, at the interface between physics and biology. One of the reasons is that, there are more and more data available, so you get closer and closer to being able to describe biology as a dynamical system and apply the physics method to biological mechanisms … even evolution,” François said.
Advances like glowing protein tags and real time imaging have led to a better understanding of cell dynamics. According to François, this is where the physics comes in.
“In physics, we are used to dynamical process …. I know if I push a ball, it’s going to roll in some way I can predict. So the question now is … can you do the same thing [with evolution], can you write a theory that will predict evolution in some way?”
A large part of François’ research is creating numerical models in hopes of finding equations that can be used to predict evolution or cellular biological processes.
“You can consider [cells] as dynamical systems that rely on computation. Like computers, they integrate a lot of external clues; and from those external clues they make decisions,” he said.
In cells, these external clues come in the form of signals from other cells or the environment. Based on these signals, decisions are made regarding whether or not to replicate, and which proteins to make. Therefore, using computers to model cell decisions is a logical step.
Recently, François used numerical models to look at the process of developing vertebrae. This is an important topic for biologists because of its medical implications; problems during vertebrate development can lead to scoliosis or trunk dwarfism.
François started with a simple question: what kind of cell processes create vertebrae? In response, he used his computer to evolve a gene network that was able to form vertebrae.
“It turned out that the network that was formed really corresponded to something we actually see in biology; and what is really striking is that, it evolves basically a genetic ‘clock,’ a genetic oscillation.”
In the genetic oscillation, waves of signals travel through the developing cells, coordinating vertebrae formation.
To test his model, François collaborated with Professor Sharon L. Amacher from Ohio State University, a biologist studying embryonic development with time lapse imaging. Amacher tested the model using genetically modified zebrafish, and found that there was an oscillating pattern during cell development.
From purely numerical models, François was able to produce the general process of vertebrae formation.
“In some way, we recapitulated what happened in evolution,” he said.
Although interdisciplinary collaborations can yield ground-breaking results, it is sometimes a challenge to communicate.
“Very often, you don’t speak the same language,” François said.
Different training and different research interests are common barriers. François has dealt with these issues by studying biology extensively, and working with biological researchers.
“By doing this you create an interaction, find a common language, and then you can do more sophisticated things.”
At McGill, one avenue for interaction is the Quantitative Biology Initiative: a multi-disciplinary research group that brings together faculty from McGill Physics, Biology, Chemistry, and Computer Science, as well as Université de Montréal researchers.
There is also a growing trend of applying physics to solve problems in other dynamic systems like economics, or even politics.
“I think this is where physics can help a lot …. Physicists are used to taking a very complicated system, and simplifying it to a core equation, a core variable, to really identify what the most important parameters are,” François said.