Code for the GPseudoClust method, including pre- and postprocessing steps.
Magdalena E Strauß, Paul DW Kirk, John E Reid, Lorenz Wernisch (2019); GPseudoClust: deconvolution of shared pseudo-trajectories at single-cell resolution.
Author of the code: ME Strauss
A tutorial is provided in .mlx and .pdf format in the files GPseudoClustByExample.mlx and GPseudoClustByExample.pdf ( inside GPseudoClust folder).
The clustering method can be run without downloading additional software.
However, for postprocessing our R package for combining PSMs using nonparametric Bayesian methods is required
library(devtools)
install_github("magStra/nonparametricSummaryPSM")
https://github.com/magStra/nonparametricSummaryPSM
For plotting we use the subaxis function by Aslak Grinsted:
https://www.mathworks.com/matlabcentral/fileexchange/3696-subaxis-subplot
The folder lmkk_summaryMatrixRepresentation contains additional methods for postprocessing, which use the following software, which requires separate download:
- Code implementing the localised kernel k-means method available at https://github.com/mehmetgonen/lmkkmeans,
Gönen, M. and Margolin, A.A. (2014). Localized data fusion for kernel k-means clustering with application to cancer biology. In Advances in Neural Information Processing Systems 27, pages 1305-1313.
-
The Mosek optimisation software (https://www.mosek.com/).
-
The SIMLR software, where one of the functions is used for the estimation of the optimal number of clusters for the summary clustering in the post-processing.
Wang, B. et al. (2017). Visualization and analysis of single-cell RNA-seq data by kernel-based similarity learning. Nat Meth, 14, 414-416.