CytoGLMM: Conditional Differential Analysis for Flow and Mass Cytometry Experiments

Abstract

BackgroundFlow and mass cytometry are important modern immunology tools for measuring expression levels of multiple proteins on single cells. The goal is to better understand the mechanisms of responses on a single cell basis by studying differential expression of proteins. We focus on cell-specific differential analysis and one fixed cell type. In contrast, most current methods learn cell types and perform differential analysis jointly. Our narrower field of application allows us to define a more specific statistical model with easier to control statistical guarantees.ResultsDifferential analysis of marker expressions can be difficult due to marker correlations and inter-individual heterogeneity, particularly for studies of human immunology. We address these challenges with two multiple regression strategies: A bootstrapped generalized linear model and a generalized linear mixed model. On simulated datasets, we compare the robustness towards marker correlations and heterogeneity of both strategies. For paired experiments, we find that both strategies maintain the target false discovery rate under medium correlations and that mixed models are statistically more powerful under the correct model specification. For unpaired experiments, our results indicate that much larger patient sample sizes are required to detect differences. We illustrate the CytoGLMMR package and workflow for both strategies on a pregnancy dataset.ConclusionsOur approach to find differential proteins in flow and mass cytometry data reduces biases arising from maker correlations and safeguards against false discoveries induced by patient heterogeneity.

Stereotypic Expansion of T Regulatory and Th17 Cells during Infancy Is Disrupted by HIV Exposure and Gut Epithelial Damage

“…Coordinates for multidimensional scaling (MDS) of markers expressed on CD4 1 cells were calculated using metaMDS function and permutational multivariate ANOVA (PERMANOVA) statistical differences performed using the adonis function from the vegan R package. Generalized linear mixed model (GLMM) analysis was performed using an open-source R package CytoGLMM as described by Seiler et al (17). Unsupervised cell population identification was performed using a self-organizing map and hierarchal clustering as implemented in the FlowSOM (Flow Self-Organizing Maps) and metaclustering R packages (18).…”

Mucosal-homing natural killer cells are associated with aging in persons living with HIV

“…Spearman correlation analysis was performed in Prism v9.0 (GraphPad Software, San Diego, California, USA). Generalized linear modeling (GLM) with bootstrapping was performed with the CytoGLMM R package 51 with R v3.6.3. 52 GLM with bootstrapping was performed according to previously published studies.…”

High-dimensional profiling of pediatric immune responses to solid organ transplantation

“…We implemented generalized linear models (GLMs) which have been commonly used in both clinical trial and single-cell analysis to compare the stable cohort to the rejection cohort, controlling for basal differences in cell type abundance between allograft types. [27][28][29] The advantage of this approach is that it allows us to leverage all samples in the same statistical test, improving statistical power. The proportions of CD8 naive, CD8 CM, and TOT cells were all significantly associated with allograft health (Figure 3C).…”