Code for reproducing the modelling presented in: H.Y.G. Lau, B.D. Fulcher, A. Fornito. Scaling of gene transcriptional gradients with brain size across mouse development, bioRxiv (2020).
This repository contains code for the modelling component; code for the main data analysis from this paper is in this repository.
Please add the support functions in the Peripheral directory by running startup.
ForSchematicAllows you to visualize some example outputs of the spatial lag model with specific parameters in an example 20 x 20 square grid:
An initial step is to get the model parameters in a good part of the space, using optimization.
The optimized values should then be set in the function GiveMeDefaultParameters: params.ensembleParams.rho and params.d0scalingFactor.
The modelling across time is done in simulateGrid.
Running this across many iterations is done in GetWithErrorBars.
For example in the paper we used 50 repeats of the model (each repeat going across the seven developmental time points) to estimate means and standard deviations of model fits across runs:
numRepeats = 50;
GetWithErrorBars(numRepeats)This saves the figure out to Model_Simo_50Reps.svg as:
