Automated Transcriptome Classifier Pipeline: Comprehensive Transcriptome Analysis
Implemented t-distributed stochastic neighbor embedding
A function to plot do a Consensus clustering to validate the results
cluster the samples
Input Expression File
Produce a Heatmap using a Supervised clustering Algorithm
A Function for Assisting Supervised Clustering
Compute Top genes
Unsupervised Clustering
An unsupervised fully-automated pipeline for transcriptome analysis or a supervised option to identify characteristic genes from predefined subclasses. We rely on the 'pamr' <http://www.bioconductor.org/packages//2.7/bioc/html/pamr.html> clustering algorithm to cluster the Data and then draw a heatmap of the clusters with the most significant genes and the least significant genes according to the 'pamr' algorithm. This way we get easy to grasp heatmaps that show us for each cluster which are the clusters most defining genes.