A Hierarchical Sheaf Spectral Embedding Framework for Single-Cell RNA-seq Analysis

View PDF HTML (experimental)

Abstract:Single-cell RNA-seq data analysis typically requires representations that capture heterogeneous local structure across multiple scales while remaining stable and interpretable. In this work, we propose a hierarchical sheaf spectral embedding (HSSE) framework that constructs informative cell-level features based on persistent sheaf Laplacian analysis. Starting from scale-dependent low-dimensional embeddings, we define cell-centered local neighborhoods at multiple resolutions. For each local neighborhood, we construct a data-driven cellular sheaf that encodes local relationships among cells. We then compute persistent sheaf Laplacians over sampled filtration intervals and extract spectral statistics that summarize the evolution of local relational structure across scales. These spectral descriptors are aggregated into a unified feature vector for each cell and can be directly used in downstream learning tasks without additional model training. We evaluate HSSE on twelve benchmark single-cell RNA-seq datasets covering diverse biological systems and data scales. Under a consistent classification protocol, HSSE achieves competitive or improved performance compared with existing multiscale and classical embedding-based methods across multiple evaluation metrics. The results demonstrate that sheaf spectral representations provide a robust and interpretable approach for single-cell RNA-seq data representation learning.

Subjects:

Machine Learning (cs.LG); Spectral Theory (math.SP); Genomics (q-bio.GN); Machine Learning (stat.ML)

MSC classes: 55N30, 62H30

Cite as: arXiv:2603.26858 [cs.LG]

(or arXiv:2603.26858v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2603.26858

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Xiang Xiang Wang [view email] [v1] Fri, 27 Mar 2026 13:52:16 UTC (135 KB)