SRFBPredictor: SRFBPredictor

In sanjaysinghrathi/SRFBPredictor: SRFBPredictor is a package to predict the reponse of a colorectal cancer patient given Chemoradiotherapy.

Description

SRFBPredictor is a package to predict the reponse of a colorectal cancer patient given Chemoradiotherapy.

Usage

SRFBPredictor(training_data_exp_matrix, training_data_phenodata,
              testing_data_exp_matrix, testing_data_phenodata,
              biomarker, combat=TRUE, mlmodel="elastic-net")

Arguments

training_data_exp_matrix	Expression matrix for training dataset having probes as rows and samples as columns
training_data_phenodata	Phenodata (clinical) data for training dataset having samples as rows and various clinical attributes as columns. Response should be first column followed by other important clinical attributes
testing_data_exp_matrix	Expression matrix for test dataset having probes as rows and samples as columns
testing_data_phenodata	Phenodata (clinical) data for test dataset having samples as rows and various clinical attributes as columns. Response should be first column followed by other important clinical attributes. Columns for test phenodata should be same as training dataset
biomarker	A character vector having gene names or probe names to be tested as signature
combat	True to adjust testing dataset gene expression values by keeping training set as reference
mlmodel	Machine learning model to predict responses ("elastic-net", "svmLinear", "svmNonLinear", "neuralNet", "randomForest")

Examples

load("SRFBPredictor.RData")

training_em <- expression_data[,1:100]
training_pData <- pData[1:100,1:3]

testing_em <- expression_data[,101:123]
testing_pData <- pData[101:123, 1:3]

biomarker <- as.character(biomarker)

SRFBPredictor(training_data_exp_matrix = training_em, training_data_phenodata = training_pData,
              testing_data_exp_matrix = testing_em, testing_data_phenodata = testing_pData,
              biomarker = biomarker)

sanjaysinghrathi/SRFBPredictor documentation built on Dec. 4, 2019, 2:08 p.m.