MPCA_CMS

In this project, a Multi-core Parallel Co-evolution Algorithm is proposed for solving the CMS model (MPCA_CMS). The input is a binary mutation matrix A, a connected PPI network Q and a parameter K. The output is a set of genes corresponding to submatrix M. In the MPCA_CMS method, due to the large number of genes and samples, it takes much time in calculating individual fitness. Therefore, the calculation of individual fitness is partitioned into a set of independent computation tasks and assigned to multi-core processors to execute simultaneously.

Operating environment：

Windows 10，R3.4.1

Input datas: a weighted non-binary mutation matrix A, a PPI network Q, a parameter K;

• binary mutation matrix: data\SNVdata_440.csv ; data\OV_SNV_2547.csv; data\SNV_332x5804.csv
  Example of A input to algorithm, Their rows represent the same set of cancer samples, and their columns represent two sets of genes.
  

• network matrix: data\second_like_data_GBM_440.csv ; data\second_like_data_2547.csv; data\second_like_data_332x5804.csv
  Example of Q file input to algorithm, Both their rows and columns represent genes.

Output: a set of genes corresponding to submatrix M;

A gene set and its corresponding fitness function value.

For example: GBM dataset, k=3, the result is: "CDKN2A" "MDM2" "TP53" , "2.85873015873016"

Steps:

1.Install and load the parallel package.

install.packages('doParallel')
library(doParallel)

2.First select the functions and run it.

3.load data

4.Initialization parameters.

5.Iterative loop