Repository files navigation

The R packageMCube implements the methods in theMMM paper.MMM, standing for theMixture of Mixed Models, is a unifiedframework for statistical identification of cell-type-specific spatiallyvariable genes in spatial transcriptomic (ST) studies.

Beginning with the raw count data,MMM uses a log-mixture structureto account for cell type composition while simultaneously correcting forthe spot and platform effects between ST and single‐cell RNA sequencing(scRNA-seq) data. The mixed-effects model decomposes thecell-type-specific gene expression in ST data into three components: theaverage gene expression of the same cell type obtained from scRNA-seqdata, spatial variations, and non-spatial variations. The statisticalsignificance of spatial variations is then examined using a powerfulnon-parametric test capable of detecting diverse spatial patterns.

You can install the development version ofMCube fromGitHub with:

# install.packages("devtools")devtools::install_github("YangLabHKUST/MCube")

The code for reproducing the real data analysis results presented in ourpaper are available on the tutorial website(https://mcube-tutorial.readthedocs.io/):

• Visium human dorsolateral prefrontal cortexdataset
• Multiple adult mouse brain datasets from differentsources
• Xenium human breast cancerdataset
• 3DDrosophila embryo model constructed from Stereo-seqdataset

If you find theMCube package or any of the source code in thisrepository useful for your work, please cite:

A unified framework for identification of cell-type-specific spatiallyvariable genes in spatial transcriptomic studies.
Zhiwei Wang, Yeqin Zeng, Ziyue Tan, Yuheng Chen, Xinrui Huang, HongyuZhao, Zhixiang Lin, and Can Yang.
2025.

The R packageMCube is developed and maintained byZhiweiWang.

Please feel free to contactZhiweiWang,Prof. HongyuZhao,Prof. ZhixiangLin, orProf. CanYang if any inquiries.