getpvalue {GraphAT} | R Documentation |
The function takes as inputs two adjacency matrices. Let X denote the observed number of edges in common between the two adjacency matrices. To test the significance of the correlation between the two data sources, the function performs N random edge permutations and random node permutations respectively. For each permutation test, the function outputs the proportion of N realizations that resulted in X edges or more at the intersection of the two datasources
getpvalue(act.mat, nonact.mat, num.iterations = 1000)
act.mat |
Adjacency matrix corresponding to first data source. That is, the i,j element of this matrix is 1 if data source one specifies a functional link between genes i and j |
nonact.mat |
Adjacency matrix corresponding to first data source. That is, the i,j element of this matrix is 1 if data source two specifies a functional link between genes i and j |
num.iterations |
Number of realizations from random edge (node) permutation to be obtained |
We note that the first adjacency matrix, denoted act.mat is the data source that is permutated with respect to edges or notes
A vector of length 2, where the first element is the P value from Random Edge Permutation and the second element is the P value from Random Node Permutation
Raji Balasubramanian
permEdgesM2M
, permNodesM2M
, makeClustM
act.mat <- matrix(0,3,3) act.mat[2,1] <- 1 act.mat[3,1] <- 1 nonact.mat <- matrix(0,3,3) nonact.mat[2,1] <- 1 nonact.mat[3,2] <- 1 p.val <- getpvalue(act.mat, nonact.mat, num.iterations = 100) print(p.val)