The anatomy of single cell mass cytometry data

Abstract

Mass cytometry enables the measurement of up to 50 features on single cell. This has catalyzed a shift toward multidimensional data analysis methods, rather than the manual gating strategies as traditionally for in flow cytometry data. This shift means that data scientists are involved in the analysis process to an increasing degree. As the data is analyzed in a more unbiased fashion, where noisy or uninformative observations are not easily excluded, a deeper knowledge of the origin, noise, and modalities of the data is therefore needed to embark on useful data analysis. In this primer, we introduce the idiosyncrasies of mass cytometry data with a focus on the technical properties of how data generated with the CyTOF® system, and the characteristics of protein expression in the cells of the hematopoietic continuum, specifically targeted toward data scientists. We also provide a comprehensive online repository of scripts, tutorials, and example data. © 2018 International Society for Advancement of Cytometry

Guidelines for the use of flow cytometry and cell sorting in immunological studies (second edition)

“…Generally, experimental workflows for mass cytometry are typically very specific to individual studies, and many factors should be considered during the setup of mass cytometry studies .…”

Relevance of Antibody Validation for Flow Cytometry

“…This allows us the consistency needed for the meaningful measurements over long periods of time (month to years) (19,20) that are primarily used for diagnostics and the monitoring of treatment efficacy (21); but they are also increasingly used in preclinical human studies (22). A requirement for the successful use of computational methods is that an assumption can be made that any changes detected in the data are related to the biological question (and are not due to technical variation) (8,27). A requirement for the successful use of computational methods is that an assumption can be made that any changes detected in the data are related to the biological question (and are not due to technical variation) (8,27).…”

“…The large data sets that result from these types of immuno-monitoring applications require a method of computational analysis for the resultant large cytometry data sets (7,(23)(24)(25)(26). A requirement for the successful use of computational methods is that an assumption can be made that any changes detected in the data are related to the biological question (and are not due to technical variation) (8,27). As with any complex interconnected system, a large-scale cytometry experiment is only as good as its weakest component.…”

ISSUE HIGHLIGHTS ‐ March 2020

“…Mass cytometry allows for the analysis of cell heterogeneity on a very large scale, theoretically measuring up to 100 different cell features simultaneously with metal isotope‐tagged antibodies without any spectral overlap (Olsen et al, ). Such a huge array of analytical data requires however special softwares for the mapping of high‐dimensional data and interpretation of the results (Amir et al, 2013, Levine et al, ; Bagwell et al, ), sometimes causing variations in the final results due to the features of their respective mathematical architecture (Amir et al, ; Levine et al, ; Li et al, ; Olsen et al, ).…”