Assessment of Batch Normalization Methods on Mass Cytometry Data
Poster Highlights
- In this study, we used 68 PBMC samples from melanoma patients to assess the effectiveness of two batch normalization tools, CytoNorm and cyCombine.
- We found that both methods preserved biological variance, through the evaluation of their MAD scores.
- CyCombine demonstrated superior performance in reducing technical variation with an Earth Mover's Distance reduction of 0.70, compared to CytoNorm's 0.37.
- The findings highlight the complexity of removing batch variation without distorting true biological signals, emphasizing the need for careful selection of normalization methods in multi-batch experiments.
Chronister WD, Leavitt LS, Black GS, et al. Assessment of batch normalization methods on mass cytometry data. 2024. The 39th International Clinical Cytometry Society Meeting.
📄 Download the PDF