SVM2CRM: support vector machine for cis-regulatory elements detections


[Up] [Top]

Documentation for package ‘SVM2CRM’ version 1.16.0

Help Pages

cisREfindbed Create and preprocessing of the input for SVM2CRM analysis
createBed Create bed file of predictions using svm.
createSVMinput Take the output of findFeatureOverlap and then create a positive and negative set of cis-regulatory elements
featSelectionWithKmeans This function select the most meaningful variables in a matrix of ChIP-seq data using k-means and ICRR.
findFeatureOverlap Find the overlap of genomic regions between the output of cisREfind and a database of validated cis-regulatory elements.
frequencyHM frequencyHM
getSignal Model the signals of each histone marks around genomic features (e.g. enhancers, not_enhancers).
performanceSVM Estimate the performance of prediction.
plotFscore Plot the F-score in relation with the sensitivity and specificity
plotROC Plot the ROC curve of the best model
predictionGW Perform prediction of cis-regulatory elements genome-wide
smoothInputFS Smooth the signals of the histone marks to prepare the input for feature selection analysis
tuningParametersCombROC Test different models using different kernel of SVM, values of cost functions, the number of histone marks.