DoIntegPPI: Integration of gene expression matrix and PPI network
In ChenWeiyan/LandSCENT: Landscape Single Cell Entropy

Description Usage Arguments Value References Examples

This function finds the common genes between the scRNA-Seq data matrix and the genes present in the PPI network, and constructs the maximally connected subnetwork and reduced expression matrix for the computation of signaling entropy.

1	DoIntegPPI(exp.m, ppiA.m, log_trans = FALSE)

`exp.m`	Can be three major kinds of input: One is a scRNA-Seq data matrix with rows labeling genes and columns labeling single cells. And it can be either a log-transformed data matrix with minimal value around 0.1 (recommended), or an nonlog-transformed data matrix with minimal value 0. The other two kinds of input can be either a "SingleCellExperiment" class object or a "CellDataSet" class object
`ppiA.m`	The adjacency matrix of a user-given PPI network with rownames and colnames labeling genes (same gene identifier as in `exp.m`)
`log_trans`	A logical. Whether to do log-transformation on the input data matrix or not. Default is FALSE

A list of two or four objects:

expMC Reduced expression matrix with genes in the maximally connected subnetwork

adjMC Adjacency matrix of the maximally connected subnetwork

data.sce/data.cds Orginal input sce/cds data objects

data Normalized data matrix

degree.v The nodes(gene) degree in the integrated network

dgC.m An additional dgCMatrix object of the input data, for the convenience of later on calculation

Chen, Weiyan, et al. Single-cell landscape in mammary epithelium reveals bipotent-like cells associated with breast cancer risk and outcome. Communications Biology 2 (2019): 306. doi: 10.1038/s42003-019-0554-8.

Teschendorff AE, Tariq Enver. Single-cell entropy for accurate estimation of differentiation potency from a cell’s transcriptome. Nature communications 8 (2017): 15599. doi: 10.1038/ncomms15599.

Teschendorff AE, Banerji CR, Severini S, Kuehn R, Sollich P. Increased signaling entropy in cancer requires the scale-free property of protein interaction networks. Scientific reports 5 (2015): 9646. doi: 10.1038/srep09646.

Banerji, Christopher RS, et al. Intra-tumour signalling entropy determines clinical outcome in breast and lung cancer. PLoS computational biology 11.3 (2015): e1004115. doi: 10.1371/journal.pcbi.1004115.

Teschendorff, Andrew E., Peter Sollich, and Reimer Kuehn. Signalling entropy: A novel network-theoretical framework for systems analysis and interpretation of functional omic data. Methods 67.3 (2014): 282-293. doi: 10.1016/j.ymeth.2014.03.013.

Banerji, Christopher RS, et al. Cellular network entropy as the energy potential in Waddington's differentiation landscape. Scientific reports 3 (2013): 3039. doi: 10.1038/srep03039.

### define a small network
ppiA.m <- matrix(0,nrow=10,ncol=10);
ppiA.m[1,] <- c(0,1,1,1,1);
for(r in 2:nrow(ppiA.m)){
  ppiA.m[r,1] <- 1;
}
rownames(ppiA.m) <- paste("G",1:10,sep="");
colnames(ppiA.m) <- paste("G",1:10,sep="");

### define a positively valued expression matrix (20 genes x 10 samples)
exp.m <- matrix(rpois(20*10,8),nrow=20,ncol=10);
colnames(exp.m) <- paste("S",1:10,sep="");
rownames(exp.m) <- paste("G",1:20,sep="");

### run integration function
Integration.l <- DoIntegPPI(exp.m,ppiA.m);
print(dim(Integration.l$expMC));
print(dim(Integration.l$adjMC));