Disentangling Shared and Target-Enriched Topics via Background-Contrastive Non-negative Matrix Factorization

2026-02-25

AI Summary

This work addresses the challenge of isolating condition-specific signals in high-dimensional biological data, which are often obscured by unknown shared confounders such as baseline biological structures or technical noise. To overcome the limitations of conventional dimensionality reduction methods, the authors propose Background-Contrastive Non-negative Matrix Factorization (BC-NMF), which introduces contrastive learning into the NMF framework for the first time. By jointly factorizing target and matched background data under a shared non-negative basis, BC-NMF explicitly disentangles common and target-specific components, yielding interpretable, condition-enriched latent variables. The method integrates GPU-accelerated multiplicative updates with mini-batch training, ensuring scalability and computational efficiency. Applied across diverse biological datasets, BC-NMF successfully uncovers masked signals—including depression-associated single-cell transcriptional programs, genotype-linked protein expression patterns in mice, leukemia treatment-specific alterations, and TP53-dependent drug responses.

Technology Category

• Machine Learning: Matrix & Tensor Methods
• Natural Language Processing: Learning & Optimization for NLP
• Data Mining & Knowledge Management: Scalability, Parallel & Distributed Systems

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Biological signals of interest in high-dimensional data are often masked by dominant variation shared across conditions. This variation, arising from baseline biological structure or technical effects, can prevent standard dimensionality reduction methods from resolving condition-specific structure. The challenge is that these confounding topics are often unknown and mixed with biological signals. Existing background correction methods are either unscalable to high dimensions or not interpretable. We introduce background contrastive Non-negative Matrix Factorization (\model), which extracts target-enriched latent topics by jointly factorizing a target dataset and a matched background using shared non-negative bases under a contrastive objective that suppresses background-expressed structure. This approach yields non-negative components that are directly interpretable at the feature level, and explicitly isolates target-specific variation. \model is learned by an efficient multiplicative update algorithm via matrix multiplication such that it is highly efficient on GPU hardware and scalable to big data via minibatch training akin to deep learning approach. Across simulations and diverse biological datasets, \model reveals signals obscured by conventional methods, including disease-associated programs in postmortem depressive brain single-cell RNA-seq, genotype-linked protein expression patterns in mice, treatment-specific transcriptional changes in leukemia, and TP53-dependent drug responses in cancer cell lines.

Problem

• disentanglement
• confounding variation
• high-dimensional data
• condition-specific signals
• background correction

Innovation

• background-contrastive NMF
• target-enriched topics
• non-negative matrix factorization
• interpretable latent components
• scalable dimensionality reduction