TreeStar’s FlowJo – Wikipedia remains a mainstay but its days of dominance may be numbered.

The most consequential changes in Flow cytometry – Wikipedia over the past decade or so were conversion from hardware to digital compensation, adoption of the logicle scale (1), and the introduction of Mass cytometry – Wikipedia. While FlowJo and its ilk could take the first two in their stride, newer data analysis approaches are necessary as it gets ever easier to assess more cellular parameters simultaneously.

After all, the ‘old’ way of analyzing flow cytometry data has considerable drawbacks.

  • Subjective manual gating (sequential partitioning of cell populations) during analysis.
    • Multi-center studies show it’s one of the most important causes of experimental variability (2).
    • Fosters confirmation bias.
  • Hierarchical analysis. Multi-dimensional data is already artificially squished into 2 dimensions (2D) by these old approaches, with associated loss of information. Imposing artificial hierarchy just adds insult to that injury.
  • Technological improvements increase the number of markers that can be assessed in one go. For example, mass cytometry is capable of simultaneously assessing ~40 parameters per cell. The conventional approach of analyzing flow cytometry data across two dimensions at a time just won’t cut the mustard.

All this to say that flow cytometry is presently going through perhaps its most consequential transition phase in decades as it starts to routinely handle previously unimagined complexity. Being in a period of flux makes it foolhardy to judge any of the newer options best. While bioinformaticians have developed several open source computational approaches, many

  • Aren’t user-friendly and require end-users with some programming knowledge.
  • Are published in specialist bioinformatics journals with little or no outreach to the immunology community.

Thus, a bunch of such programs are floating around and some get tested by some niche users. However, getting one or more of them into mainstream use requires the hard work of running them through their paces and validating them through multi-center field trials, which means comparing their output to that of the current norm. That entails considerable effort.

In recent years, the FlowCAP (Flow Cytometry: Critical Assessment of Population Identification Methods) consortium has made considerable strides in validating some open source automated gating algorithms. In a 2016 paper (3), they found open source OpenCyto (http://opencyto.org/) performed adequately in a head-to-head comparison with FlowJo. Substantial advantages of approaches such as OpenCyto are

  • Replacing subjective manual with more objective automated gating.
  • Better mitigation of hierarchical analysis.

Bioconductor (Bioconductor – Home) is the best known repository of such open source flow cytometry data analysis software. It has ~42 packages (4) for handling various aspects of flow cytometry data. Some recent papers (5, 6, 7) discuss pros and cons of the relatively more widely used R, MatLab or Python-based packages such as Citrus (nolanlab/citrus, 8), Phenograph (Dana Pe’er Lab, 9), SPADE (10), t-SNE (11), Wanderlust (12).


1. Tirumalai Kamala’s answer to What tools do researchers in the life sciences commonly use to perform basic statistical analyses of their data?

2. Maecker, Holden T., J. Philip McCoy, and Robert Nussenblatt. “Standardizing immunophenotyping for the human immunology project.” Nature Reviews Immunology 12.3 (2012): 191-200. https://www.researchgate.net/pro…

3. Finak, Greg, et al. “Standardizing flow cytometry immunophenotyping analysis from the human immunophenotyping consortium.” Scientific reports 6 (2016): 20686. http://www.nature.com/articles/s…

4. Bioconductor – BiocViews

5. Mair, Florian, et al. “The end of gating? An introduction to automated analysis of high dimensional cytometry data.” European journal of immunology 46.1 (2016): 34-43. http://onlinelibrary.wiley.com/d…

6. Newell, Evan W., and Yang Cheng. “Mass cytometry: blessed with the curse of dimensionality.” Nature Immunology 17.8 (2016): 890-895.

7. Saeys, Yvan, Sofie Van Gassen, and Bart N. Lambrecht. “Computational flow cytometry: helping to make sense of high-dimensional immunology data.” Nature Reviews Immunology 16.7 (2016): 449-462.

8. Bruggner, Robert V., et al. “Automated identification of stratifying signatures in cellular subpopulations.” Proceedings of the National Academy of Sciences 111.26 (2014): E2770-E2777. http://www.pnas.org/content/111/…

9. Levine, Jacob H., et al. “Data-driven phenotypic dissection of AML reveals progenitor-like cells that correlate with prognosis.” Cell 162.1 (2015): 184-197. https://pdfs.semanticscholar.org…

10. Qiu, Peng, et al. “Extracting a cellular hierarchy from high-dimensional cytometry data with SPADE.” Nature biotechnology 29.10 (2011): 886-891.

11. Amir, El-ad David, et al. “viSNE enables visualization of high dimensional single-cell data and reveals phenotypic heterogeneity of leukemia.” Nature biotechnology 31.6 (2013): 545-552. https://pdfs.semanticscholar.org…

12. Bendall, Sean C., et al. “Single-cell trajectory detection uncovers progression and regulatory coordination in human B cell development.” Cell 157.3 (2014): 714-725. http://ac.els-cdn.com/S009286741…