Ab initio spillover compensation in mass cytometry data

Abstract

Signal intensity measured in a mass cytometry (CyTOF) channel can often be affected by the neighboring channels due to technological limitations. Such signal artifacts are known as spillover effects and can substantially limit the accuracy of cell population clustering. Current approaches reduce these effects by using additional beads for normalization purposes known as single-stained controls. While effective in compensating for spillover effects, incorporating single-stained controls can be costly and require customized panel design. This is especially evident when executing largescale immune profiling studies. We present a novel statistical method, named CytoSpill that independently quantifies and compensates the spillover effects in CyTOF data without requiring the use of single-stained controls. Our method utilizes knowledge-guided modeling and statistical techniques, such as finite mixture modeling and sequential quadratic programming, to achieve optimal error correction. We evaluated our method using five publicly available CyTOF datasets obtained from human peripheral blood mononuclear cells (PBMCs), C57BL/6J mouse bone marrow, healthy human bone marrow, chronic lymphocytic leukemia patient, and healthy human cord blood samples. In the PBMCs with known ground truth, our method achieved comparable results to experiments that incorporated single-stained controls. In datasets without ground-truth, our method not only reduced spillover on likely affected markers, but also led to the discovery of potentially novel subpopulations expressing functionally meaningful, cluster-specific markers. CytoSpill (developed in R) will greatly enhance the execution of large-scale cellular profiling of tumor immune microenvironment, development of novel immunotherapy, and the discovery of immune-specific biomarkers. The implementation of our method can be found at https://github.com/KChen-lab/CytoSpill.git.

Full spectrum flow cytometry and mass cytometry: A 32‐marker panel comparison

“…However, recent publications have reported that signal spillover due to issues such as metal oxidation, abundancy related spillover, and metal impurities can impact data quality and interpretation in MC [ 52 , 69 ]. Efforts have been employed using single metal bead based controls [ 52 ], unlabeled competitor antibody [ 69 ], or statistical approaches [ 70 ] for compensating crosstalk between channels to improve overall data quality.…”

Dysbiosis of a microbiota–immune metasystem in critical illness is associated with nosocomial infections

“…Normalized mass cytometry data files were further processed in R using the CytoSpill package v.0.1.0 to correct for any signal overlap between markers 66 . Next, corrected FCS files were imported into Cytobank (Cytobank) for manual gating on CD45 + single-cell events and major cell populations (Supplementary Table 8 ).…”

Spatial omics technologies at multimodal and single cell/subcellular level

“…As some spatial platforms gather information on a per spot basis, several denoising software tools such as content-aware image restoration (CARE) [72], residual channel attention networks (RCAN) [73], and Noise2Void [74] can be applied to raw images to improve the quality especially in fluorescence based platforms. For mass cytometry based spatial platforms, data quality can be improved by spillover correction software [75]. For example, if the ground truth of spillover is known or empirically adjusted from the data, Reinforcement Dynamic Spillover EliminAtion (REDSEA) [76] leverages the spatial proximity of cells with signal that comes from generally mutually exclusive markers.…”