bioRFR: A MATLAB(R) Package for biological System Resilience Function Reconstruction
by Yan Li (linay0124@outlook.com), Shaowu Zhang* (zhangsw@nwpu.edu.cn)
Introduction
In this repository, we provide a MATLAB(R) package for reconstructing resilience function of biological processes from gene expression data and the code about biological analysis and figures plotting associated with the paper: Resilience function uncovers the critical transitions in cancer initiation.
Data Available
The data used in our paper is available at Google Drive. And the detail information can be found in README file.
Directory Tree
├─Algorithm [the code of biRFR framework]
| ├─Parameter_calculate.m [the function used to calculate status and resilience parameters]
| ├─Estimate_landscape.m [the function used to estimate potential landscape]
| ├─Resilience_centrality.m [the function used to calculate resilience centrality for each genes]
| ├─computeGrid.m [a internal function called by Estimate_landscape()]
| └─example.m [a sample script illustrates the workflow of bioRFR]
├─BiologicalAnalysis [the code of biological analysis and figures plotting]
| ├─PlotResilienceFunction.R [plotting the resilience function shown in Fig.3 in main text of paper]
| ├─PredictSurvivalTime.R [analyzing the relationship between status parameter and survival time (Fig. 4)]
| ├─SurvivalAnalysis_KGs.R [survival Analysis based on KGs (Fig.5)]
| └─TopologicalDistance.R [calculating topological distance between iKGs and CGC genes]
├─Data [the detail information about available data]
How to use bioRFR
We provide a sample script Algorithm/example.m which contains the complete workflow of bioRFR. In brief, it consists with the following steps.
1. Data preparation
bioRFR needs two gene expression matrix whose rows and columns represent genes and samples respectively. One of them contains reference samples <exp_ref> and the other consists with the remaining samples <exp_other>.
2. Calculating status and resilience parameters
One can call the following function to calculate parameters for every samples:
[x_effN,beta_effN,x_effT,beta_effT] = Parameter_calculate(<exp_ref>,<exp_other>,th);3. Estimation potential landscape
Based on the output of step 2, the potential landscape and corresponding minima and maxima can be estimate using Estimate_landscape() function.
[minlocs,maxlocs] = Estimate_landscape(x_effN,beta_effN,x_effT,beta_effT,peakth);4. Calculation resilience centrality of each gene for every samples
In order to assess the importance of each gene for state transition of individual patient, the resilience centrality of each gene can be calculated using Resilience_centrality() function.
[resN_beta,resT_beta] = Resilience_centrality(expN,expT,th);5. Downstream biological analysis
Based on resilience function and personalized key genes (KGs), one can perform downstream biological analysis (see BiologicalAnalysis\ as some examples).