Part 3.  Data Analysis

CyTOF data generally requires some pre-processing, for example normalization for signal intensity and debarcoding of individual samples from a composite file.  Review of files for technical quality is also a good idea, either by an automated method or manual gating.  Manual gating of CyTOF data is most commonly done using FlowJo or Cytobank.  However, it is difficult if not impossible to visualize all aspects of 40+ parameter data using manual gating; so automated clustering and visualization tools have become common.  Resources below review the most common platforms and algorithms, and also point to public sources of data.

An overview of CyTOF data analysis (Kimball and al., J.Immunol., 2018)

An extensive review (Olsen and al., Journal of Quantitative Cell Science, 2018)

Data pre-processing resources:  The Clambey lab has linked guides to signal intensity normalization, debarcoding, and gating on live intact singlets.  Bead normalization and data transformations are also reviewed in the biosurf tutorial.

Basic gating in FlowJo:  A tutorial on gating in FlowJo can be found or through this tutorial (from Boston Children Flowlab).


Cytobank gating and automated analysis:  Cytobank has a collection of videos on basic gating as well use of their embedded automated algorithms. In the Cytobank support page, you can find a detailed summary of how to do the analysis.

Further readings on data analysis:

The anatomy of single cell mass cytometry data (Olsen et al. , Journal of Quantitative Cell Science, 2018)

A comparison of CyTOF analysis methods (Weber and Robinson, Cytometry 2016)

Related: A data scientist’s primer to analysis of mass cytometry data

A beginner’s guide to analyzing and visualizing mass cytometry data (Kimball et al., Journal of Immunology, 2018)

Minimizing batch effects in mass cytometry data (Schuyler et al., Frontiers in Immunology, 2019)

Acquisition, processing and quality control of mass cytometry data (Lee et al., Mass Cytometry, 2019)

Standardizing immunophenotyping data for the human immunology project (Maecker et al., Nature Reviews, 2012)

A Universal Live Cell Barcoding-Platform for Multiplexed Human Single Cell Analysis (Hartmann et al.,  Scientific Reports, 2018)


Quick links to publications of some CyTOF data analysis tools:

SPADE (Qiu et al., Nature Biotechnology, 2011)

viSNE (Amir et al., Nature Biotechnology, 2013)

Citrus (Bruggner et al., PNAS, 2014)

FlowSOM (Van Gassen et al., Journal of Quantitative Cell Science, 2015)

UMAP (Becht et al., Nature Biotechnology, 2018)


Public data repositories:  Depositing flow and CyTOF data in the public domain has become more common, and is increasingly being required by funding agencies and/or journals.  These databases in turn become resources for those who would like to re-use data for new purposes.  An introduction presentation introducing these repositories can be found here.

Major repositories that can accept CyTOF data include:

•       Flow Repository:  Flow Repository has the advantage that it is designed like Cytobank, specifically for flow or CyTOF data.  This makes uploading and downloading data and annotations easy, especially for those who already use Cytobank.

•       ImmPort:  ImmPort is a more comprehensive data warehouse, designed to accommodate many data types, including but not limited to flow cytometry and CyTOF data.  It has relatively strict requirements for data formatting and annotation.  An introductory video can be found on their website.  ImmPort Galaxy is a web tool for exploration of data through this introductory video below.