Automated identification of stratifying signatures in cellular subpopulations

Abstract

Elucidation and examination of cellular subpopulations that display condition-specific behavior can play a critical contributory role in understanding disease mechanism, as well as provide a focal point for development of diagnostic criteria linking such a mechanism to clinical prognosis. Despite recent advancements in singlecell measurement technologies, the identification of relevant cell subsets through manual efforts remains standard practice. As new technologies such as mass cytometry increase the parameterization of single-cell measurements, the scalability and subjectivity inherent in manual analyses slows both analysis and progress. We therefore developed Citrus (cluster identification, characterization, and regression), a data-driven approach for the identification of stratifying subpopulations in multidimensional cytometry datasets. The methodology of Citrus is demonstrated through the identification of known and unexpected pathway responses in a dataset of stimulated peripheral blood mononuclear cells measured by mass cytometry. Additionally, the performance of Citrus is compared with that of existing methods through the analysis of several publicly available datasets. As the complexity of flow cytometry datasets continues to increase, methods such as Citrus will be needed to aid investigators in the performance of unbiased-and potentially more thorough-correlation-based mining and inspection of cell subsets nested within high-dimensional datasets.informatics | biomarker discovery

Identification of enhanced IFN-γ signaling in polyarticular juvenile idiopathic arthritis with mass cytometry

“…Citrus, a computational technique combining hierarchical clustering with an analysis of stratifying differences in cluster features (in our case, phosphorylation of a panel of signaling proteins within specific immune cells) between two groups of samples, was performed with the R "citrus" package on fcs files gated on live immune cells to compare treatment-naive patients with matched controls (or remission samples) under different stimulation conditions (45). Surface markers were clustering parameters.…”

Modeling of inter‐sample variation in flow cytometric data with the joint clustering and matching procedure

“…The performance of the JCM procedure is compared with some other available methods for automated clustering and sample classification, including HDPGMM (6), flowMatch (27), Citrus (17), and ASPIRE (8). Like FLAME, JCM is based on finite mixtures of skew distributions.…”

“…In particular, many of these methods proceed by training a classifier such as a support vector machine (SVM) to discriminate the samples between different classes. Some examples include flowBin (3,14), SWIFT (3,15), ASPIRE (8,16), PBSC (3), Citrus (17), and flowPeaks (3,18) which train a classifier based on features derived from a fitted model or a clustering of the data. Typically, the cluster proportions (size) are used as the feature vector.…”

Nanotechnologies for biomedical science and translational medicine

“…For example, single-cell analysis is increasingly applied toward understanding immune responses triggered during various cancer immunotherapies (22)(23)(24)(25) or for unraveling the functional behaviors that emerge from heterogeneous healthy or diseased tissues (26,27). Aside from flow-and mass-cytometry methods (28,29), virtually every single-cell -omics platform involves some level of microfluidics and/or nanotech. Specifically, a single cell has only a certain number of copies of any given analyte.…”