At the Biological Complexity: From Molecules to Systems meeting, Perdita Stevens gave an excellent talk on work that she is doing towards modelling viral infection. The two viruses that she spoke about were influenza and the Semliki Forest Virus – the latter spends part of its life-cycle in mosquitos.
Simulating the intracellular behaviour of the virus is ever so complicated – the process is made up of many equations each requiring that you input numbers: How many viruses should enter the cell? What is the rate of binding of the viral nucleic acid to the relevant polymerase? What are the rates of all the steps towards making the capsid proteins and then assembling these around the nucleic acid? What is the cell doing about stopping the virus from doing all this?
In many cases, the numbers are not actually known and either experiments need to be done to get the “real numbers” or you need to take a guess and see what happens (whether the model behaves in a way similar to “reality”).
This problem, reminded me of a game that we used to play at university called WA-TOR. It looked like this:
and simulated the dynamics of shark and fish populations that had the specifications that you plugged in.
It has since evolved somewhat (as things do) and now looks like this:
(green = fish, red = sharks, blue = water)
Anyway, Prof. Stevens is using DIZZY (an open source tool developed by Stephen Ramsey) for the virus simulations. I don’t pretend to understand such things, but the quote below from the user manual seems to imply that it is useful to think of the system-to-simulate in terms of a series of “chemical equations”. This pleases me!
“Dizzy is a chemical kinetics simulation software package implemented in Java. It provides a model definition environment and various simulation engines for evolving a dynamical model from specified initial data. ... A model consists of a system of interacting chemical species, and the reactions through which they interact. The software can then be used to simulate the reaction kinetics of the system of interacting species.”
Other things from the talk that were pleasing (and that I want to find out more about) were:- that there seems to be something different about cell death in mosquitos in comparison to mammals
- mosquitos seem to only have an “innate” immune system, and RNAi plays an important part in it
- the wild-type variant of the Semlicki Forest Virus is gone by day 16 after infection and is replaced by a small-plaque, attenuated variant.
Its surprising how such a simple model can produce apparently realistic biological behaviour. This model appears to be as simple as:
if (Shark.x = Fish.x) and (Shark.y = Fish.x)
then EatFish(Even_If_Ate_Recently, Yum)
if (Shark.LastTimeAte = SharkStarve)
then KillShark(Rest_In_Peace)
if (Shark.Age = SharkBreed)
then NewShark(Another_Shark_Not_Required)
if (Fish.Age = FishBreed)
then NewFish(Another_Fish_Not_Required)
DrunkWalk(All_Remaining_Sharks)
DrunkWalk(All_Remaining_Fish)
DoNextIteration()
A slightly more complex model might, say, introduce some food for the fish and so on.
I recall some work I did a while back where some complex (and not well understood) chemical reactions were modelled just by taking the k constant from a book for the reaction between two species. With enough imagination (basically guessing what intermediate and final products can be produced) and using exactly the same “finite time slice” approach used in Wa-Tor, you can get reasonably accurate results.
While it would be easy to sneer at these very crude theoretical approaches, they are tremendous fun and an intellectual challenge for the theoreticians and computer programmers! Ocasionally, they’re very practical indeed…
I think you mean: Shark.y = Fish.y … ?