terrysun0302/FastMix: An OLS-based approach for fitting LMER with applications to gene expression deconvolution analysis

This package implements a computationally efficient method for fitting linear mixed model. It is based on the OLS fitting and moment-matching and is 30 ~ 60 times faster than the standard LMER fitting algorithm (e.g., \code{lme4}. One application of this method is to combine the gene expression deconvolution analysis and downstream differential expression analysis into one step. To this end, we include a hypothesis testing framework to identity significant interactions between the clinical variables, gene expressions, and cell types. In addition, we implemented several discriminant scores based the FastMix model, so that the users may apply a wide range of machine learning algorithms on these scores to build predictive models.

Getting started

Package details

AuthorHao Sun, Xing Qiu
MaintainerHao Sun <terrysun0302@gmail.com>
Package repositoryView on GitHub
Installation Install the latest version of this package by entering the following in R:
terrysun0302/FastMix documentation built on Nov. 14, 2019, 4:54 a.m.