How can we better analyze and interpret the complex data generated by high-dimensional cytometry technologies in biological research?
@NatureComms #UniversityofBonn
"Unveiling the power of high-dimensional cytometry data with cyCONDOR"
• High-dimensional cytometry (HDC) has become widely available but lacks adequate analytical methods to fully utilize its complex data, with current tools either being web-hosted with limited scalability or designed for expert computational biologists, creating a need for more accessible and comprehensive analysis solutions.
• The authors introduce cyCONDOR, an integrated computational framework that provides guided preprocessing, clustering, dimensionality reduction, and machine learning algorithms to analyze HDC data from different technologies like CyTOF, HDFC, SpectralFlow, and CITE-seq, which can measure up to 50 markers simultaneously at single-cell resolution.
• The framework implementation includes batch correction with Harmony, pseudotime analysis with slingshot, kNN classifier for data projection, and CytoDx for clinical classification, tested on various datasets including PBMCs from HIV patients showing >99% accuracy for cell type prediction and perfect classification of 10 test samples.
• The results demonstrated cyCONDOR's ability to handle millions of cells while running on consumer hardware, with validation across multiple datasets showing successful batch correction, trajectory inference of HSC differentiation, and clinical sample classification with high accuracy (>90% for AML detection), providing both basic analysis and advanced interpretation features.
#cytometry #bioinformatics #singlecell #Immunology
Authors: Charlotte Kröger, Sophie Müller, Jacqueline Leidner et. al
@LorenzoBonaguro
Link:
nature.com/articles/s41467-0…