rfTrain: rfTrain

In Babulab-bioc/MSiP: 'MassSpectrometry' Interaction Prediction

Description

The labeled feature matrix can be used as input for Random Forest (RF) classifier. The classifier then assigns each bait-prey pair a confidence score, indicating the level of support for that pair of proteins to interact. Hyperparameter optimization can also be performed to select a set of parameters that maximizes the model's performance. This function also computes the areas under the precision-recall (PR) and ROC curve to evaluate the performance of the classifier.

Usage

rfTrain(
  dtInput,
  impute = TRUE,
  p = 0.3,
  parameterTuning = TRUE,
  mtry = seq(from = 1, to = 10, by = 2),
  min_node_size = seq(from = 1, to = 9, by = 2),
  splitrule = c("gini"),
  metric = "Accuracy",
  resampling.method = "repeatedcv",
  iter = 5,
  repeats = 5,
  pr.plot = TRUE,
  roc.plot = TRUE
)

Arguments

dtInput	Data frame containing instances with class labels
impute	Logical value, indicating whether to impute missing values
p	The percentage of data that goes to training; defaults to 0.3
parameterTuning	Logical value; indicating whether to tune rf hyper parameters
mtry	Number of variables to possibly split at in each node and it is bound by the number of variables in your model
min_node_size	Minimal node size
splitrule	Splitrule rule for classification: 'gini', 'extratrees' or 'hellinger' with default 'gini'
metric	A string that specifies what summary metric will be used to select the optimal model; default to Accuracy
resampling.method	The resampling method:'boot', 'boot632', 'optimism_boot', 'boot_all', 'cv', 'repeatedcv', 'LOOCV', 'LGOCV'; defaults to repeatedcv
iter	Number of resampling iterations; defaults to 5
repeats	for repeated k-fold cross validation only; defaults to 5
pr.plot	Logical value, indicating whether to plot precision-recall (PR) curve
roc.plot	Logical value, indicating whether to plot ROC curve

Value

Data frame containing a classification results for all instances in the data set, where positive confidence score corresponds to the level of support for the pair of proteins to be true positive, whereas negative score corresponds to the level of support for the pair of proteins to be true negative.

Author(s)

Matineh Rahmatbakhsh, [e-mail]

Examples

data(testdfClassifier)

predidcted_RF <-
rfTrain(testdfClassifier,impute = FALSE, p = 0.3, parameterTuning = FALSE,
mtry  = seq(from = 1, to = 5, by = 1),
min_node_size = seq(from = 1, to = 5, by = 1),
splitrule =c("gini"),metric = "Accuracy",
resampling.method = "cv",iter = 2,repeats = 2,
pr.plot = TRUE, roc.plot = FALSE)
head(predidcted_RF)

Babulab-bioc/MSiP documentation built on Dec. 17, 2021, 9:52 a.m.