Version 2.6.0 o Major: We fixed a bug in DaMiR.ModelSelect. Now optimal models are correctly selected; o Major: Now users can plot specific graphs in DaMiR.Allplot and we added new plots; o Minor: We modified the color scale in corrplot ---------------- Version 2.0.0 o Since version 2.0.0 of the software, DaMiRseq offers a solution to solve two distinct problems, in supervised learning analysis: (i) finding a small set of robust features, and (ii) building the most reliable model to predict new samples; o The functions DaMiR.EnsembleLearning2cl_Training, EnsembleLearning2cl_Test and EnsembleLearning2cl_Predict were deprecated and replaced by DaMiR.EnsL_Train, DaMiR.EnsL_Test and DaMiR.EnsL_Predict, respectively; o We have created a new function DaMiR.ModelSelect to select the best model in a machine learning analysis; o We have created two new functions DaMiR.iTSnorm and DaMiR.iTSadjust to normalize and adjust the gene espression of independent test sets; o Two types of expression value distribution plot were added to the DaMiR.Allplot function. ---------------- Version 1.5.2 o The DaMiR.normalization function embeds also the 'logcpm' normalization. o Now, DaMiR.EnsembleLearning calculates also the Positive Predicted Values (PPV) and the Negative Predicted Values (NPV). o Three new functions have been implemented for the binary classification task: DaMiR.EnsembleLearning2cl_Training, DaMiR.EnsembleLearning2cl_Test and DaMiR.EnsembleLearning2cl_Predict. The first one allows the user to implement the training task and to select the model with the highest accuracy or the average accuracy; the second function allows the user to test the selected classification model on a test set defined by the user; the last function allows the user to predict the class of new samples. o Removed black dots in the violin plots. ---------------- Version 1.4.1 o Adjusted Sensitivity and Specificity calculations. ---------------- Version 1.4 Relevant modifications: o DaMiRseq performs both binary and multi-class classification analysis. o The 'Stacking' meta-learner can be composed by the user, setting the new parameter 'cl_type' of the DaMiR.EnsembleLearning function. Any combination up to 8 classifiers ('RF', 'NB', 'kNN', 'SVM', 'LDA', 'LR', 'NN', 'PLS') is now allowed. o If the dataset is imbalanced, a 'Down-Sampling' strategy is automatically applied. o The DaMiR.FSelect function has the new argument, called nPlsIter, which allows the user to have a more robust features set. In fact, several feature sets are generated by the 'bve_pl' function, setting 'nPLSIter' parameter greater than 1. Finally, an intersection among all the feature sets is performed to return those features which constantly occur in all runs. However, by default, nPlsIter = 1. Minor modifications and bugs fixed: o DaMiR.Allplot accepts also 'matrix' objects. o The DaMiR.normalization function estimates the dispersion, through the parameter 'nFitType'. o In the DaMiR.normalization function, the gene filtering is disabled if 'minCount = 0'. o In the DaMiR.EnsembleLearning function, the method for implementing the Logistic Regression has been changed to allow multi-class comparisons; instead of the native 'lm' function, the 'bayesglm' method is now used. o The new parameter 'second.var' of the 'DaMiR.SV' function, allows the user to take into account a secondary variable of interest (factorial or numerical) that the user does not wish to correct for, during the sv identification. ---------------- Version 1.3.7 o DaMiRseq performs multi-class classification analysis. o The Stacking meta-learner can be composed by the user, setting the new parameter 'cl_type' of the DaMiR.EnsembleLearning() function. Any combination of the 8 classifiers is now allowed. o If the dataset is imbalanced, a 'Down-Sampling' strategy is automatically applied. o The DaMiR.FSelect() function has the new argument, called 'nPlsIter', which allows the user to have a more robust features set. In fact, several feature sets are generated by the bve_pls() fuction (embedded in DaMiR.FSelect()), setting 'nPLSIter' parameter greater than 1. Finally, an intersection among all the feature sets is performed to return those features which constantly occur in all runs. However, by default, 'nPlsIter = 1'. o DaMiR.Allplot() accepts also 'matrix' objects as well as NA values (which are not plotted). o The DaMiR.normalization() function estimates the dispersion, through the parameter 'nFitType'; as in DESeq2 package, the argument can be 'parametric' (default), 'local' and 'mean'. o In the DaMiR.normalization() function, the gene filtering is desabled if 'minCount = 0'. o In the DaMiR.EnsembleLearning() function, the method for implementing the Logistic Regression has been changed to allow multi-class comparisons; instead of the native 'lm' function, 'bayesglm' method implemented in the caret 'train' function, properly set, is now used. o The new parameter 'second.var' of the DaMiR.SV() function, allows the user to take into account a secondary variable of interest (factorial or numerical) that the user does not wish to correct for, during the sv identification.