MLSeq Machine Learning Package for RNA-Seq Data

This document is a technical guide for the MLSeq package, which applies machine learning algorithms to classify next-generation RNA-Sequencing (RNA-Seq) data. It outlines the functionalities of MLSeq, including the ability to preprocess RNA-Seq count data, apply various machine learning models, and evaluate their performance. The guide details the preprocessing methods available, such as normalization techniques like DESeq median ratio and TMM normalization, as well as transformations like log-cpm and variance stabilizing transformation. Users can build classification models using over 90 available algorithms, including both microarray-based classifiers and novel RNA-Seq specific methods. The document also explains how to prepare input data, split datasets into training and test sets, and optimize model parameters. Additionally, it provides instructions on predicting class labels and comparing classifier performances. The guide aims to assist researchers in effectively utilizing the MLSeq package for their RNA-Seq classification tasks.