Automated flow cytometric analysis across large numbers of samples and cell types

Abstract

Multi-parametric flow cytometry is a key technology for characterization of immune cell phenotypes. However, robust high-dimensional post-analytic strategies for automated data analysis in large numbers of donors are still lacking. Here, we report a computational pipeline, called FlowGM, which minimizes operator input, is insensitive to compensation settings, and can be adapted to different analytic panels. A Gaussian Mixture Model (GMM)-based approach was utilized for initial clustering, with the number of clusters determined using Bayesian Information Criterion. Meta-clustering in a reference donor permitted automated identification of 24 cell types across four panels. Cluster labels were integrated into FCS files, thus permitting comparisons to manual gating. Cell numbers and coefficient of variation (CV) were similar between FlowGM and conventional gating for lymphocyte populations, but notably FlowGM provided improved discrimination of "hard-to-gate" monocyte and dendritic cell (DC) subsets. FlowGM thus provides rapid high-dimensional analysis of cell phenotypes and is amenable to cohort studies.

Methods for discovery and characterization of cell subsets in high dimensional mass cytometry data

“…Unfortunately, expert manual gating has been shown to be particularly prone to inter-operator variability [7] and a tendency to overlook cell populations [8–10]. Recent efforts have developed new tools for high dimensional cytometry data that bring in elements of machine learning and statistical analysis, including clustering [11–14], dimensionality reduction [8], variance maximization [15], mixture modeling [6, 16–18], spectral clustering [19], neural networks [20], and density-based automated gating [21]. Here, we highlight use of these tools in a sequential single cell bioinformatics workflow (Table 1).…”

A macro-DAG structure based mixture model

“…[4,8] demonstrate the practical advantage of using the Gaussian mixture model relative to the traditional gating approach. It is shown that the performances of this automated approach can compete with manual expert analyses, and furthermore has the added benefits of speed, objectivity, reproducibility, and the ability to evaluate several subpopulations in many dimensions simultaneously [9]. The mixture model approach was recently improved in order to deal with large data sets and compared with the top-ranked approaches [1].…”

Automated Flow Cytometric MRD Assessment in Childhood Acute B‐ Lymphoblastic Leukemia Using Supervised Machine Learning

“…Methods such as flowMeans (25), flowMerge (26), or FLOCK (27) identify cell populations in FCM data by means of unsupervised hard clustering. In the context of automated FCM analysis, Gaussian Mixture Models (GMMs) have been successfully applied (23,26,(29)(30)(31). In the context of automated FCM analysis, Gaussian Mixture Models (GMMs) have been successfully applied (23,26,(29)(30)(31).…”

“…In the context of automated FCM analysis, Gaussian Mixture Models (GMMs) have been successfully applied (23,26,(29)(30)(31). There are only a few supervised approaches for automated FCM analysis (30,(32)(33)(34). There are only a few supervised approaches for automated FCM analysis (30,(32)(33)(34).…”

“…Other authors such as Shapiro et al (28) and Naim et al (29) propose using soft decisions for clustering. In the context of automated FCM analysis, Gaussian Mixture Models (GMMs) have been successfully applied (23,26,(29)(30)(31). As a parametric density model GMMs form a compact representation of complex non-Gaussian distributions, advanced approaches such as SWIFT (29) and the Hierarchical Dirichlet Process Gaussian Mixture Model (31) extend conventional GMMs with the focus on detecting rare cell populations.…”