Jinghan Huang

The paper develops an in-silico cell sorting method to identify granulocyte-specific gene expression biomarkers from bulk blood data, enabling discrimination between bacterial and viral infections.

This paper introduces CoDA-hd, a compositional data analysis approach for scRNA-seq, showing its advantages in dimension reduction, clustering, and trajectory inference compared to log-normalization.

This paper develops a ratio-based biomarker (DIRECT LS-TA) using monocyte-specific gene expression ratios in peripheral blood to improve viral infection triage, outperforming the IFN score in diagnostic accuracy.

The paper shows that trans-differentiation of plasmablasts to neutrophils in scRNA-seq data is an artefact caused by low-quality cells and highlights the need for improved statistical standards and quality control in single-cell analysis.

This paper develops an imputation algorithm for single-cell RNA-seq and benchmarks it against other methods across various downstream analyses.