SimulationData: Simulate y and X from a given network structure.

In xueweic/APGD: TGPred: Efficient methods for predicting target genes of a transcription factor by integrating statistics, machine learning, and optimization.

Simulate y and X from a given network structure.

Usage

SimulationData(N_samples, N_genes, Adj, Sigma, method, beta0=NULL, beta_true=NULL)

Arguments

N_samples	the number of sample size.
N_genes	the number of traget genes.
Adj	the adjacency matrix of network structure. Adjacency matrix must be a N_genes*N_genes dimensional symmetric matrix, the elements equal 1 indicates two genes are connected. If you consider Barabasi-Albert Network or Hierarchical Network in the article, you can directly use "ConstructNetwork" function to get the adjacency matrix.
Sigma	the covariance matrix of target genes according to network structure. You can directly use "GraphicalModel" function to get the covariance matrix.
method	"HN": by Hierarchical Network, "BAN": by Barabasi-Albert Network or "DIY": by user designed
beta0	numeric value of effect size in simulation settings. # default: NULL; if method is "HN" or "BAN", input a nunerical value.
beta_true	numeric matrix with the dimension of N_genes * 1 in simulation settings. # default: NULL; if method is "DIY", input a nunerical matrix (N_genes * 1).

Value

y	expression levels of a transcription factor (TF)
X	expression levels of n_genes target genes (TGs)
beta	true regulated effect beta for N_genes TGs.

Examples

N_samples <- 300
    
N_genes <- 200
    
Adj = ConstructNetwork(N_genes, "BAN")
    
Sigma1 = GraphicalModel(Adj)

    # Set up a true regression coefficient for simulated data (beta0=1)
res = SimulationData(N_samples,N_genes,Adj,Sigma1,"BAN", beta0 = 1)
    
y = res$y
    
X = res$X
    
beta1 = res$beta

xueweic/APGD documentation built on Sept. 4, 2023, 2:18 a.m.