build_network: build_network constructs weighted neighborhood network

In Zouter/MPath: Mpath: a single-cell RNAseq data analysis algorithm that maps multi-branching single-cell trajectories revealing cell progression during development

Description

build_network constructs weighted neighborhood network

Usage

build_network(exprs, landmark_cluster, distMethod = "euclidean",
  baseName = NULL, writeRes = TRUE)

Arguments

exprs	a data frame or matrix of expression data(ie. rpkm, TPM, fpkm) containing cells in columns and genes in rows
landmark_cluster	a data frame or matrix of two columns or a tab delimited file of landmark cluster assignment of individual cells. The first column indicates cell ID, the second column indicates the landmark cluster which the cell was assigned to.
distMethod	the method for calculating dissimilarity between cells. distMethod can be one of "pearson", "kendall", "spearman" or "euclidean". Default is "euclidean".
baseName	output directory
writeRes	a boolean to indicate whether to save result files

Value

a matrix of weighted neighborhood network, column and row names are landmarks, the values represent the weights of the edges connecting two landmarks

Examples

## Not run: 
exprs = "TPM_GSE60783_noOutlier_geneQC0.05anyGroup_CD4vsCD8DEG.txt";
baseName = "TPM_GSE60783_noOutlier_geneQC0.05anyGroup_CD4vsCD8DEG";
landmark = "TPM_GSE60783_noOutlier_geneQC0.05anyGroup_CD4vsCD8DEG_landmark_cluster.txt"; 
# or landmark can be the return value of landmark_designation function 
neighbor_network <- build_network(exprs = exprs,
                                  landmark_cluster = landmark,
                                  baseName = baseName)

## End(Not run)

Zouter/MPath documentation built on May 4, 2019, 2:31 p.m.