Kidney Week

Abstract: TH-PO002

Advancing Precision Medicine with scSpectra: Single-Cell Functional Profiling of Individual Patients with Kidney Diseases

Session Information

• Augmented Intelligence for Prediction and Image Analysis
  October 24, 2024 | Location: Exhibit Hall, Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Augmented Intelligence, Digital Health, and Data Science

• 300 Augmented Intelligence, Digital Health, and Data Science

Authors

• Kloetzer, Konstantin A., University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States

Konstantin A. Kloetzer, MD

• Abedini, Amin, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States

Amin Abedini, MD

• Balzer, Michael S., University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States

Michael S. Balzer, MD

• Liang, Xiujie, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States

Xiujie Liang, MD, PhD

• Levinsohn, Jonathan, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States

Jonathan Levinsohn, MD, PhD

• Ha, Eunji, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States

Eunji Ha, PhD

• Schuller, Max, Medizinische Universitat Graz, Graz, Steiermark, Austria

Max Schuller, MD

• Dumoulin, Bernhard, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States

Bernhard Dumoulin, MD

• Hogan, Jonathan, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States

Jonathan Hogan, MD

• Eller, Kathrin, Medizinische Universitat Graz, Graz, Steiermark, Austria

Kathrin Eller, MD

• Bloom, Roy D., University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States

Roy D. Bloom, MD, MBChB

• Zhang, Nancy, University of Pennsylvania Wharton School, Philadelphia, Pennsylvania, United States

Nancy Zhang, PhD

• Susztak, Katalin, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States

Katalin Susztak, MD, PhD

Background

Single-cell RNA sequencing atlases, comprising millions of cells from hundreds of individuals, could guide molecular precision diagnostics in nephrology. However, the clinical application of single-cell transcriptomics to identify cell-type-specific molecular changes in individual patients remains unestablished.

Methods

We developed scSpectra, a novel computational tool that quantifies changes in gene expression coordination across cellular functions in individual samples. To demonstrate scSpectra's capabilities, we created the largest known human single-nuclei atlas. We expanded our cross-species integrated single-cell kidney atlas to include 150 kidney samples and biopsies from various diseases, such as diabetic kidney disease, chronic kidney disease with hypertension, acute kidney injury, and ADPKD. Comprising over 700,000 cells, our atlas, combined with scSpectra, powers our Single-Cell Functional Profiling Report, identifying the most significantly dyscoordinated pathways in each cell type of individual patients.

Results

By analyzing 100 patients, we demonstrate scSpectra's ability to distinguish diseases based on cell type involvement. For instance, a patient with IgA nephropathy showed the most significant changes in inflammation-related functions within podocytes, which was rarely observed in other diseases. Comparing the dyscoordination prevalence of thousands of cellular functions across diseases, we identified functions frequently dyscoordinated in CKD patients with hypertension but not in those without hypertension. We also found disease-specific changes for ADPKD samples, sex differences, and heterogeneity in potentially druggable pathways such as Interleukin-1-related signaling.

Conclusion

With scSpectra, we introduce one of the first computational frameworks for functional contextualization and precision diagnostics from scRNA-seq data. Based on a novel statistical approach, scSpectra revolutionizes single-cell analysis and enables a detailed examination of individual samples. Implemented in clinical investigations, scSpectra could identify patients most likely to benefit from specific treatments, advancing precision medicine in the management of complex diseases.

Funding

• NIDDK Support