Abstract

Mass cytometry(Cytometry by Time-Of-Flight, CyTOF) is a single-cell technology that is ableto quantify multiplex biomarker expressions and is commonly used in basic lifescience and translational research. However, the widely used Gadolinium(Gd)-based contrast agents (GBCAs) in magnetic resonance imaging (MRI) scanningin clinical practice can lead to signal contamination on the Gd channels in theCyTOF analysis. This Gd contamination greatly affects the characterization ofthe real signal from Gd-isotope-conjugated antibodies, severely impairing the CyTOFdata quality and ruining downstream single-cell data interpretation.

To overcome thisdrawback, we developed a computational method named by GdClean for removing theGd contamination induced by the usage of GBCAs. We first confirmed that theintensities of Gd signals in the Gd contamination data were highly correlatedin pairs, and their pairwise intensity ratios were well accordant with theratios of their natural abundances. Using the control CyTOF dataset of a samplestained with specifically designed antibody panels, we identified the linear compositionof the observed Gd signals from Gd-labeled mAbs and GBCAs-induced Gdcontamination. Based on that, we designed an optimization-based method toestimate and further remove the Gd contamination at the single-cell level. Usingthe real and simulated contamination CyTOF datasets, we demonstrated that our GdCleanmethod can effectively remove the Gd contamination signal and well preserve thereal signal from Gd-labeled mAbs on single cells, demonstrating its capabilityto ensure high data quality in CyTOF data preprocessing for better downstreamsingle-cell analysis.

Diagram of Gd contaminationsignal source and GdClean algorithm design