ap-browning / scratchibm

Supplementary material for "Identifying density-dependent interactions in collective cell behaviour"

scratchIBM

Supplementary data, and code, for

"Browning AP, Jin W, Plank MJ, Simpson MJ (2019) Identifying density-dependent interactions in collective cell behaviour" J R Soc Interface 17:20200143.

Preprint available on bioRxiv.

Experimental data

Raw data in the /Data folder corresponds to experiment indices in the main document as follows:

Index	Folder
1	8000_E2
2	8000_G2
3	10000_H1
4	10000_H2
5	8000_F2
6	10000_A2
7	12000_H3
8	12000_F3
9	12000_D2

Experimental data is recorded 12 h after seeding. Hence, raw data indicated by filename ending corresponds to the following times:

Filename end	Time
"_12h"	0 h
"_30h"	18 h
"_48h"	36 h

For more details, see ExperimentalData.xlsx

Code

Requirements

• MATLAB R2018a or later
• MEX and C compiler
• (optional) Parallel computing toolbox If you do not have this toolbox, you must replace parfor with for in the following files:
  1. ABC_SimulateModel.m
  2. ABC_SMC.m

Installation

Before use, you must compile the MEX code contained in IBM.c using the Install.m script or typing:

mex IBM.c  -R2018a

into the command window.

Examples

Simulating IBM

See Example_IBM.m for an example IBM realisation.

ABC Rejection

• Parallel computing toolbox highly recommend
• This function can take a large amount of time to run for Nsamples > 100
  • We have set Nsamples = 100 and alpha = 0.1 for demonstrative purposes

Run Example_ABCRejection.m code (in the main document, we use Nsamples = 100000)

ABC SMC Model Selection

• Parallel computing toolbox highly recommend
• This function can take a large amount of time to run for N > 10
  • We have set N = 8 and not used the full sequence of thresholds for demonstrative purposes

Run Example_ABCSMC.m code (in the main document, we use Nsamples = 5000)