Biological Scaling Theory

Overview

This repository uses methods described in the following paper: http://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1004455

The paper details four methods used to calculating scaling exponents of vascular networks:

1. A conservation based method
2. A ratio based method
3. A regression based method
4. A distribution based method

This repository contains MATLAB code to generate plots to compare data outputted by two versions of vessel extraction software: Angicart (written in OCAML) and Angicart++ (written in C++)

The datasets used in this repository are for academic research purposes and can be found here: https://drive.google.com/drive/folders/1o7fbUU70ivLRVF-NcohQUgnZu-PkPM9U?usp=sharing

• files labeled with a _c extension are outputs of the C++ data (ie. patient1_c.txt)

Running the Code

Notes: The directory with data files must be added to your project path. Make sure the MATLAB Statistics and Machine Learning toolbox is installed

1. Run calcAll.m to generate all plots for a single dataset (change cName for the filename of the C++ data and aName for the filename of the Angicart data)
2. compareLen.m and compareRad.m generate histograms to compare the lengths of vessels and the radii of vessels
3. ratioBased.m, conservationBased.m, regressionBased.m, and distributionBased.m generate plots using each of the methods for both length and radii of a given dataset
4. Helper functions: importOption.m, fillArr.m, getBinFreq.m, calcA.m

Contribution

If you would like to contribute to this project, visit https://vsavage.faculty.biomath.ucla.edu/Code/HTML/indexangicplusplus.html