classification_results: Example of output from designSampleSizeClassification...

In MSstatsSampleSize: Simulation tool for optimal design of high-dimensional MS-based proteomics experiment

Description

It is the output of designSampleSizeClassification function with a list of simulated_datasets generated under same protein number and sample size. The list should include the required elements as below.

Usage

classification_results

Format

A list with five elements

Details

• num_proteins : the number of simulated proteins

• num_samples : a vector with the number of simulated samples in each condition

• results : a list with ‘num_proteins’ elements. Each element has (1) classification models trained on each simulated dataset; (2) the predictive accuracy on the validation set predicted by the corresponding classification model.

• mean_predictive_accuracy : the mean predictive accuracy over all the simulated datasets.

• mean_feature_importance : the mean protein importance vector over all the simulated datasets, the length of which is ‘num_proteins’.

• predictive_accuracy : a vector of predictive accuracy on each simulated dataset.

• feature_importance : a matrix of feature importance, where rows are proteins and columns are simulated datasets. the length of which is ‘num_proteins’.

Examples

classification_results$num_proteins
classification_results$num_samples
classification_results$mean_predictive_accuracy
head(classification_results$mean_feature_importance)

MSstatsSampleSize documentation built on Nov. 8, 2020, 4:53 p.m.