Efficient Imputation for Patch-based Missing Single-cell Data via Cluster-regularized Optimal Transport

2026-01-21

AI Summary

Single-cell sequencing data often exhibit large-scale structured missingness, posing significant challenges for conventional imputation methods that struggle to balance accuracy and computational efficiency. To address this issue, this work proposes CROT, a novel algorithm that, for the first time, integrates clustering-based regularization into an optimal transport framework. This approach effectively captures the underlying missing-data structure in high-dimensional heterogeneous single-cell datasets. By leveraging the geometric properties of optimal transport and incorporating cluster-level information, CROT achieves high imputation accuracy while substantially reducing computational overhead. Extensive experiments demonstrate that the method scales efficiently to large-scale single-cell data, offering both practical utility and strong performance in real-world applications.

Technology Category

• Search and Optimization: Combinatorial Optimization
• Computer Vision: Learning & Optimization for CV
• Machine Learning: Optimization

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Security and Privacy: Data transparency and provenance

Abstract

Missing data in single-cell sequencing datasets poses significant challenges for extracting meaningful biological insights. However, existing imputation approaches, which often assume uniformity and data completeness, struggle to address cases with large patches of missing data. In this paper, we present CROT, an optimal transport-based imputation algorithm designed to handle patch-based missing data in tabular formats. Our approach effectively captures the underlying data structure in the presence of significant missingness. Notably, it achieves superior imputation accuracy while significantly reducing runtime, demonstrating its scalability and efficiency for large-scale datasets. This work introduces a robust solution for imputation in heterogeneous, high-dimensional datasets with structured data absence, addressing critical challenges in both biological and clinical data analysis. Our code is available at Anomalous Github.

Problem

• single-cell sequencing
• missing data
• patch-based missing
• data imputation
• structured data absence

Innovation

• optimal transport
• cluster regularization
• patch-based missing data
• single-cell imputation
• scalable algorithm