Further MNN algorithm development


This repository contains code for further development of the mutual nearest neighbours (MNN) batch correction method, as implemented in the mnnCorrect and fastMNN functions in the batchelor package. It is based on the code at https://github.com/MarioniLab/MNN2017, which accompanies the paper Batch effects in single-cell RNA-sequencing data are corrected by matching mutual nearest neighbors by Haghverdi et al. (2018).

More details on practical usage can be found in the documentation for each function, as well as in the OSCA book - specifically in the Data Integration chapter and several workflows like the merged pancreas workflow.


To run the simulations, enter the simulations/ directory and run:

Real data

Three real data analyses are available - haematopoiesis, pancreas and droplet. Each subdirectory will usually contain:

Data file downloads are performed using the BiocFileCache package to save time and bandwidth. This will cache the files locally after the initial download, and reuse the cached versions when the script is re-run.