ClusterDE: a post-clustering differential expression (DE) method robust to false-positive inflation caused by double...

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Standard single-cell RNA sequencing (scRNA-seq) analysis involves clustering for cell type identification followed by differential expression (DE) testing.
• This common approach suffers from
• double dipping
• , where the same data is used for both clustering and DE gene identification, inflating false positives for cell-type marker genes.

Approach:

• We introduce ClusterDE, a novel post-clustering DE method designed to control the false discovery rate (FDR) irrespective of clustering accuracy.
• ClusterDE employs a unique strategy of generating real-data-based synthetic null data with a single cluster to rigorously evaluate the entire clustering and DE testing pipeline.

Key Points:

• ClusterDE provides robust FDR control, effectively mitigating false positives in DE gene identification.
• The method successfully identifies genuine cell-type marker genes and differentiates them from housekeeping genes.
• ClusterDE is computationally efficient, transparent, and compatible with various clustering algorithms and DE tests.

Conclusions:

• ClusterDE offers a reliable solution for accurate cell-type marker gene discovery in scRNA-seq data analysis.
• Its applicability extends beyond scRNA-seq to other post-clustering DE analyses, including single-cell multi-omics data.