Science is what we know. It’s described in encyclopedias and books and journals and is used every day. But the act of science is discovering new knowledge. How do we do that? Can we just turn the handle and another nugget of knowledge pops out? Sadly no. We don’t know how to find new science — we don’t even know what there is to know! But we do have a method to work our way through a discovery.
The scientific method is question, theorise, predict, test, analyse. For biology, biologists run the experiments. Often a statistician is drafted in for the analysis. I am a modeler; my step is prediction. I take a hypothesis and turn it into a model. The model runs, producing simulations. These simulations predict the system behaviour.
A prediction gives something to compare our experiments to. Spending time on this step helps us construct better experiments, ones more likely to show what we’re interested in. With modeling we can also play: try variations on our hypothesis, and see the effect. The step of prediction bridges a hypothesis to an experiment. The first step is to confirm that the hypothesis is coherent and complete. By producing a model, I guarantee that: otherwise the model could not work.
Recent tradition in science is to specialise individuals in an area of science. This gives a narrow focus: silos. To counter this, we can specialise each step of the method, yielding a broad, interdisciplinary team. My day job is to apply my skills to biology research. For development as a specialist, I have two focuses: internal and external. Internal is to improve my skills: find how to predict better, more reliably. I must extend my range with more model types, more systems covered. External is my communication with the other actors in the method. I develop diagramming techniques and define documentation.
I have worked on several areas of biology now. I started with developmental patterning, such as stripe formation on fish. Next I looked at enzyme chemistry, the everyday dynamic of cell biology. Returning to developmental biology, I examined the physical reorganization of muscle cells. I produced computer models of rounding, sorting and migration of cells. Most recently, I tried to understand cancer. I produced cell population models: cancer phenotype, treatment, phylogeny. My models are of cells interacting: I give the cells a few simple rules and set them going. The systems that emerge seem much more complex than the rules!
Take a look: edwardflach.org