infer_network: Infer network

In diogocamacho/nsmblR: Network inference ensemble modeler

Description

This function infers a gene expression network based on pre-defined methods.

Usage

infer_network(method = c(1, 2, 3, 4, 5, 6, 7), data, quantile_thr)

Arguments

method	Integer defining the method to be used. Defaults to using CLR (method 3).
data	Gene expression data frame or mutual information matrix
quantile_thr	Quantile threshold for edge assignment (defaults to 0.95)

Details

Inference methods:

1. Spearman correlations

2. Partial correlations with information theory (PCIT)

3. Context Likelihood of Relatedness (CLR)

4. ARACNe

5. MRNET

6. MRNETB

7. Mutual Rank (MutRank)

Value

An inferred network matrix

Examples

# Run CLR algorithm:
D <- nsmblR::data_matrix
N <- infer_network(method = 3, data = D, quantile_thr = 0.95)

# Run MutRank algorithm:
D <- nsmblR::data_matrix
N <- infer_network(method = 7, data = D, quantile_thr = 0.95)

# Clean data and then run the ARACNE algorithm:
D <- nsmblR::data_matrix
cd <- data_cleanup(D)
D <- D[-cd$nix_rows, ]
genes <- nsmblR::genes[-cd$nix_rows, ]
N <- infer_network(method = 4, data = D, quantile_thr = 0.8)

diogocamacho/nsmblR documentation built on Aug. 5, 2020, 1:20 a.m.