Lasso: miRNA target prediction with the Lasso method

In miRLAB: Dry lab for exploring miRNA-mRNA relationships

Description

Calculate the Lasso regression coefficient of each pair of miRNA-mRNA, and return a matrix of coefficients with columns are miRNAs and rows are mRNAs.

Usage

Lasso(datacsv, cause, effect, targetbinding = NA)

Arguments

datacsv	the input dataset in csv format
cause	the column range that specifies the causes (miRNAs), e.g. 1:35
effect	the column range that specifies the effects (mRNAs), e.g. 36:2000
targetbinding	the putative target, e.g. "TargetScan.csv". If targetbinding is not specified, only expression data is used. If targetbinding is specified, the prediction results using expression data with be intersected with the interactions in the target binding file.

Value

A matrix that includes the Lasso regression coefficients. Columns are miRNAs, rows are mRNAs.

References

1. Le, T.D., Zhang, J., Liu, L., and Li, J. (2015) Ensemble Methods for miRNA Target Prediction from Expression Data, submitted.

2. Tibshirani, R. (1996). Regression shrinkage and selection via the lasso. J. R. Stat. Soc. Series B Stat. Methodol., 267-288.

Examples

dataset=system.file("extdata", "ToyEMT.csv", package="miRLAB")
results=Lasso(dataset, 1:3, 4:18) 

miRLAB documentation built on Nov. 8, 2020, 5:45 p.m.