digeR {digeR}R Documentation

digeR GUI

Description

Start the Graphical User Interface for digeR.
digeR supports spots correlation analysis, score plot, classification, feature selection and power analysis.

Usage

digeR()

Details

digeR GUI options:

File Read in data and image, quit
Open upload the txt file
Upload_gel_image upload the JPG image as a reference for spots correlation analysis
Quit dispose the GUI
——————–
Correlation Spots correlation analysis
Dataset select the group to look at
Spot List select the spot to look at
Selected feature upload the feature list from feature selection
Load features upload the feature list from an saved R workspace
Pearson, Kendall, Spearman type of correlation coefficient to be calculated: "pearson" (default), "kendall", or "spearman"
Show the correlation plot the spots with required correlation
Correlation Coefficiency change the coefficiency threshold
Show spot ID plot spots with ID
Show number Show ID for those spots with required correlation
——————–
Score Plot PCA and PLSR score plot
Plot Type select either PCA or PLSR score plot
Top N component plot score plot with top N components
Pair-wise plot selected 2 components
Component 1 and 2 two components in the pairwise plot
Group set the color for the two groups
With label plot the sample ID
Scaling scale the data before plotting
Plot plot the score plot
——————–
Classification Classification
Methods select the method for the classification
Scaling scale the data before classification
Arguments
Method way for estimate the covariance matrix.
"moment" standard estimators of the mean and variance
"mle" MLEs,
"mve" to use cov.mve
"t" robust estimates based on a t distribution
nComp number of component for fitting PCR or PLSR
N-fold CV number of fold in the cross validation
nboot number of bootstrap in the classification
Selected feature upload the feature list from feature selection
Load features upload the feature list from an saved R workspace
leave-one-out cv classification with leave-one-out cross validation
N-fold cv classification with n-fold cross validation
Bootstrap classification with bootstrap
Run classification press button to do the classification
Save save the prediction results into an R workspace
Legened where the legend will be put
ROC curve generate ROC plot
Prediction result store the classification results in the selected items
——————–
Feature Selection Select important features
Method select feature selection method
Scaling scale the data before feature selection
Arguments
Method same as Method in Classification
Ncomp same as ncomp in Classification
Top select the top n variables from the feature selection
Ntree number of trees to grow in randomForest
Mtry Number of variables randomly sampled as candidates at each split. Default sqrt(number of variables)
Mfinal the number of iterations for which boosting is run or the number of trees to use
Run feature selection press to start feature selection
Select featuers store the selected features in the selected items
Save features save the features into an R workspace
——————–
Power Power analysis
Single Spots univariate power analysis
Gel multivariate power analysis for experiment design
Significant level set the significant level
Power set the power level to be achieved
Sample size per group sample size for achieving certain significant level and power in each group
Spot Number set the spots to calculated
Calculate calculate the one being left blank (either power, sample size or significant level)

Note

digeR is built upon gWidgets package. Make sure gWidgets package is properly installed.

Author(s)

Yue Fan yue.fan@ucd.ie,Thomas Brendan Murphy, R. William G. Watson


[Package digeR version 1.2 Index]