Consensus Clustering of Single Cell Transcriptomics Reveals Novel Cell Types in Pediatric Leukemia

• Israa M Shamkh

  Chief Computational Chemistry Department EBO Bio Solution Company, London, United Kingdom

  Author

• Ahmed Hassen Shntaif

  Assist Professor of Chemistry Department of Chemistry, University of Babylon, Alhilla, Iraq

  Author

• Mohammed Al-Majidi

  Department of Chemistry, College of Science for Women, University of Babylon, Alhilla 51002, Iraq

  Author

• Aziz-Ur Rehman

  Keystone Pharmacogenomics LLC, Bensalem, PA, United States

  Author

• Ihosvany Camps

  Full Professor. Department of Physics, Federal University of Alfenas, Alfenas, MG, Brazil

  Author

To apply consensus clustering to integrate single cell RNA-seq data from pediatric patients with B-cell precursor acute lymphoblastic leukemia (B-ALL) and identify robust and reproducible cellular subgroups. Advancements in single cell transcriptomics have enabled high-resolution characterization of cellular heterogeneity in complex diseases like cancer. However, analysis of single cell data remains challenging due to technical noise and biological variations between samples. Clustering algorithms used for cell type identification can be sensitive to these factors, resulting in inconsistent and irreproducible cluster definitions across datasets. To demonstrate the effectiveness of consensus clustering in overcoming technical noise and inter-patient biological variation in single cell transcriptomics data, and to reveal novel cellular subpopulations in B-ALL that are recurrent but often missed or obfuscated in standard analyses. We applied consensus clustering, an ensemble clustering approach, to integrate single cell RNA-seq data from 55 pediatric patients with B-ALL. Consensus clustering combines results from multiple independent clustering of the data to identify robust and reproducible cell populations. We compared consensus clusters to standard clustering results on individual samples and validated the biomarkers and cellular properties of the consensus subgroups through immunophenotyping.

Consensus clustering identified 22 robust cellular subgroups across patients, several of which did not correspond to standard B-ALL subtypes. The novel subgroups showed distinct gene expression profiles and phenotypic differences from known B-ALL cells. Some represented minor populations that were frequently missed or merged with major clusters in single sample analyses due to noise and heterogeneity. Our results demonstrate that consensus clustering is effective for overcoming technical noise and inter-patient biological variation in single cell transcriptomics data. It reveals novel cellular subpopulations in B-ALL that are recurrent but often missed or obfuscated in standard analyses. Integrating results from multiple clustering allows identification of robust and reproducible cell types across complex single cell datasets.

• PDF

• Biotechnology Section
• Computational Science Section

This work is licensed under a Creative Commons Attribution 4.0 International License.