Abstract
OBJECTIVES: A recent single-cell RNA sequencing study by Wilk et al. suggested that plasmablasts can transdifferentiate into 'developing neutrophils' in patients with severe COVID-19 disease. We explore the evidence for this. METHODS: We downloaded the original data and code used by the authors in their study to replicate their findings and explore the possibility that regressing out variables may have led the authors to overfit their data. RESULTS: The lineage relationship between plasmablasts and developing neutrophils breaks down when key features are not regressed out, and the data are not overfitted during the analysis. CONCLUSION: Plasmablasts do not transdifferentiate into developing neutrophils. The single-cell RNA sequencing is a powerful technique for biological discovery and hypothesis generation. However, caution should be exercised in the bioinformatic analysis and interpretation of the data and findings cross-validated by orthogonal techniques.