Single-cell RNA sequencing for cell cycle factor validation
A research team needed help analysing single-cell RNA sequencing data to validate cell cycle-related factors, investigate a bimodal mitochondrial gene expression pattern, and handle sparse data.
The challenge
A team investigating cellular mechanisms had difficulty computationally analysing complex scRNA-seq data. They needed to validate hypotheses about cell cycle-related factors, examine a bimodal distribution in mitochondrial gene expression, and address the technical challenges posed by data sparsity.
What we set out to do
- Validate cell cycle-related factors in single-cell data
- Investigate the biological significance of the mitochondrial bimodality
- Address technical challenges in sparse scRNA-seq data
- Produce publication-quality visualisations
- Provide biological interpretations of all identified factors
Our approach
Data architecture analysis
Applied pseudo-bulking to address data sparsity, used strategic data filtering protocols, mapped technical and biological sources of variation, and built a reproducible analytical workflow.
Technical implementation
Processed aggregated counts with DESeq2 using variance stabilising transformation, performed dimensionality reduction with PCA and UMAP, ran differential expression analysis between cell clusters, and applied quality control to identify technical artefacts.
Solution delivery
Delivered Jupyter notebooks with reproducible code, publication-quality visualisations, biological interpretation of identified factors, and recommendations for downstream analyses.
Results and impact
The analysis clearly identified distinct cell cycle phase groupings and validated differential expression at a 0.1% FDR threshold. Cluster 17 was identified as a technical artefact rather than a biological phenomenon, and enrichment of cell cycle pathways was confirmed among differentially expressed genes. The work gave the team a clearer view of technical versus biological variation and a stronger framework for validating biological signals against noise.
“Umbizo was a pleasure to work with. Their attention to detail and high-quality work were excellent throughout.”
Timeline
- Initial assessment and planning: 1 day
- Initial analysis completion: 7 days
- Advanced analysis delivery: 9 days
- Final delivery: 11 days
Looking forward
Expanding the analytical framework to explore interactions between cell guides and cell cycle progression, developing new visualisation techniques for complex single-cell data, and supporting publication preparation with biological interpretations.
