CRITICAL ASSESSMENT OF AUTOMATED FLOW CYTOMETRY DATA ANALYSIS TECHNIQUES

By Nima Aghaeepour, Greg Finak, Holger H. Hoos, Tim R. Mosmann, Ryan Remy Brinkman, Raphael Gottardo, Richard H. Scheuermann, Michael Biehl. Nature Methods, 2013

629 citations recorded by the source corpus.

Traditional methods for flow cytometry (FCM) data processing rely on subjective manual gating. Recently, several groups have developed computational methods for identifying cell populations in multidimensional FCM data. The Flow Cytometry: Critical Assessment of Population Identification Methods (FlowCAP) challenges were established to compare the performance of these methods on two tasks: (i) mammalian cell population identification, to determine whether automated algorithms can reproduce expert manual gating and (ii) sample classification, to determine whether analysis pipelines can identify characteristics that correlate with external variables (such as clinical outcome). This analysis presents the results of the first FlowCAP challenges. Several methods performed well as compared to manual gating or external variables using statistical performance measures, which suggests that automated methods have reached a sufficient level of maturity and accuracy for reliable use in FCM data analysis.

Subfields: Molecular Biology

View Paper: https://www.semanticscholar.org/paper/bc90ecedff8c93e929aaab39781df7327c9263b0

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

The Cell Ontology 2016: enhanced content, modularization, and ontology interoperability (2016, methodology): “…order to facilitate the analysis of data generated in high-dimensional flow cytometry or CyTOF, the flowCL software package matches cell populations identified via automated gating algorithms against existing cell types in the CL based on their combinations of markers, or immunophenotypes [72, 73].”
Identification and visualization of multidimensional antigen-specific T-cell populations in polychromatic cytometry data (2015, background): “…is generally accepted in the computational flow community that no one gating algorithm or approach will solve all specific computational problems (4,16), and integration of these disparate tools into a coherent computational pipeline has generally been challenging, particularly for large data sets.”
Detection of Minimal Residual Disease in B Lymphoblastic Leukemia Using viSNE (2015, methodology): “(13) have assessed numerous automated methods for clustering subpopulations within multiparametric data sets, but these methods are insufficiently sensitive to detect MRD populations.”
FlowSOM: Using self‐organizing maps for visualization and interpretation of cytometry data (2015, methodology): “We compute this in the same way as is explained in the FlowCAP I article (11).”
Automatic Classification of Cellular Expression by Nonlinear Stochastic Embedding (ACCENSE) (2013, background): “Important advances have been made toward developing better analytic tools for multivariate cytometry data (7).”
Orchestrating single-cell analysis with Bioconductor (2019, methodology): “While many algorithms and software packages have previously been used to cluster data from bulk assays and single-cell flow cytometry [95], the complexities of scRNA-seq data pose unique challenges for clustering tasks, specifically a richer feature space, large numbers of cells, and data sparsity [”
Domain Generalization by Marginal Transfer Learning (2017, methodology): “We used a data set that is a part of the FlowCAP Challenges where the ground truth labels have been supplied by human experts (Aghaeepour et al., 2013).”
Automated Mapping of Phenotype Space with Single-Cell Data (2016, methodology): “The same comparison was repeated for SPADE as well as top-performing flow data clustering methods that participated in FlowCAP competition on the same set of data 4 .”
Computational flow cytometry: helping to make sense of high-dimensional immunology data (2016, background): “A systematic comparison of the performance of automated gating techniques can be found in Aghaeepour et al. 28 .”
Sensitive detection of rare disease-associated cell subsets via representation learning (2016, methodology): “Unsupervised approaches have been extended to the classiﬁcation of single-cell samples and have been successful where disease association manifested itself in condition-speciﬁc differences of abundant cell subpopulations 8,9 .”
Nonparametric Estimation of Multi-View Latent Variable Models (2013, methodology): “We used the DLBCL Lymphoma dataset collection from Aghaeepour et al. (2013) to compare our kernel algorithm with multi-view mixture of Gaussian model.”
Background fluorescence and spreading error are major contributors of variability in high‐dimensional flow cytometry data visualization by t‐distributed stochastic neighboring embedding (2018, background): “As a consequence, unsupervised algorithms for the identi ﬁ cation of cell populations in high-dimensional data sets would provide an objective approach for data analysis (9).”
CyTOF workflow: differential discovery in high-throughput high-dimensional cytometry datasets (2017, positive): “Considerable effort has been made to improve and automate cell population identification, such as unsupervised clustering (Aghaeepour et al., 2013).”