PEA_ml: Transcriptome-wide m5C predictor training under machine...

In cma2015/PEA-m5C: An integrated R toolkit for plant m5C analysis.

Description

This function used for transcriptome-wide m5C predictor construction. First, the fixed number (parameter "independent_num") of independent samples (positive and negative samples) are randomly sampled from training samples. The the k-fold cross-validation would be performed based on the training samples but excluding the independent samples. Finally, the m5C predictor, performance evaluation on independent test datasets and cross-validation results will be returned.

Usage

PEA_ml(pos_sample,neg_sample,independent_num=100,ig="ALL",
                   ratio = 1,modeltype = "RFC",cvnum = 5,repeatTimes = 1, ntree=200,over_sampling = F)
                   

Arguments

pos_sample	A numeric matrix recording the features for positive sample.
neg_sample	A numeric matrix recording the features for nagative sample.
independent_num	A numeric value, the number of independent sample
feature_num	A numeric value, the number of selected features based the top of information gain rank, the "ALL" means all features
modeltype	A character string, which specifies machine learing method.
cvnum	An integer value, the number of fold for cross validation.
repeatTimes	An integer value,If the negative sample is larger than the limit of the positive sample, the number of the negative samples and the number of samples of the positive sample is repeated
over_sampling	Logical value, where TRUE represents balance the positive and negative samples according to the ratio based smote simulation
ratio	A numeric value, where 1 represents balance the positive and negative sample.

Value

A list of result.

The first level is used feature num group.

The second level is cross validation group.

The third level is the detail information including positives.test.score.id, negatives.test.score.id, positives.test.score,negatives.test.score, positives.test, negatives.test, auc_test, auc_test_id

Author(s)

Jie Song, Jingjing Zhai, Chuang Ma

Examples

load(paste0(system.file(package = "PEAm5c"),"/data/samples.Rds"))
aaa <- PEA_ml(pos_sample = pos_sample,neg_sample = neg_sample)
aaa

cma2015/PEA-m5C documentation built on May 17, 2019, 8:05 a.m.