EmbedSOM is an efficient self-organizing-map-based dimensionality reduction algorithm developed for use in flow- and mass-cytometry data analysis, but applicable also as a general dimensionality reduction method similarly as tSNE or UMAP. The algorithm is described in a separate article [1] that is currently published on bioRxiv.
Its resource efficiency improves existing cytometry data analysis by delivering interactive-speed embedding, thus vastly simplifying the human-computer interaction efficiency of the workflow. The software is currently published on GitHub [2] as a R package.
A separate website [3] provides a tutorial usage and describes several related workflow tools and a GPU implementation. [1] Kratochvíl, Miroslav, et al. "SOM-based embedding improves efficiency of high-dimensional cytometry data analysis." bioRxiv (2019): 496869. https://www.biorxiv.org/content/10.1101/496869v2.full [2] https://github.com/exaexa/EmbedSOM/ [3] http://bioinfo.uochb.cas.cz/embedsom/