Model 3 - Cell Migration


Model Details

Model 3 is a Stochastic Differential Equation system that describes the behaviour of cells migrating as a response to chemotaxis, based on a pseudopod-centred mechanism. The system is defined and described in Neilson et al. (2011). The implementation of the system for Cside 2018 follows that of Tweedy et al. (2013). Luke Tweedy has very kindly written some supplementary notes for the event which detail the equations and how to use the code that solves the system. Please note there is a difference between the equations in these notes and in the publication of Neilson et al. (2011), so to be clear - competitors should follow the equations from the notes available here - Tweedy Supplementary Notes.

For the purposes of the event, we will be making the assumption that all the output is measurable in practice i.e. that we can take measurements of the global inhibitor, local inhibitor, local activator and stimulus strength. This is not likely to be true in practice and the difficulty of the inference task could be increased by working with the system with some of the variables unobserved. However, in order that the inference task is achievable in the given time frame for Cside 2018, we will assume that the variables are observable.

The units (u) for time are arbitrary and the default value for "duration" is 100,000u. Time increments with a step size of 0.1u between 0.1u and "duration" and the code samples every 100u (evenly spaced), resulting in 1000 timepoints in total. Hence, the spacing is: 0.1u, 100.1u, 200.1u, 300.1u, ... , 99900.1u. This should not be changed by the competitor. There are 10 parameters to be inferred; fa, rc, kb, db, Db, km, sa, ba, Da and da. The details of these parameters are available in the supplementary notes. 

For the creation of your own synthetic data, all output files produced by the code (global inhibitor, local activator, local inhibitor, stimulus strength, x co-ordinate and y co-ordinate of each finite element node) should have Gaussian noise added to them as detailed here.

Thank you

Thank you to Luke Tweedy and Robert Insall of the Beatson Institute for helping Cside 2018 to work with this model for this year's event. A special thanks to Luke for his code and for helping make it fit for purpose, as well as his time discussing the model, code and for writing the supplementary notes.


The code is run by using a java exectuable file. For details on how to implement the code, including how to pass arguments from the command line, please see Luke Tweedy's supplementary notes (above). In the output files of the code, the rows of the files represent different time points and the columns represent different finite element nodes.

You can download a .zip file of the code here -

You can download a .7z file of the code here - CellMigration.7z