What question did this study set out to answer?

To enhance molecular classification in cancers by developing a framework that efficiently identifies subtypes using multi-omics data.

April 21, 2026Open Access

ULSL: Unified Latent and Similarity Learning for Robust Multi-Omics Cancer Subtype Identification

Key Points

To enhance molecular classification in cancers by developing a framework that efficiently identifies subtypes using multi-omics data.
Proposed Unified Latent and Similarity Learning (ULSL) framework for simultaneous learning of latent embeddings and similarity matrices.
Utilized graph fusion to ensure structural consistency across different omics data.
Evaluated performance on synthetic datasets and ten public cancer datasets from The Cancer Genome Atlas.
ULSL outperformed seven state-of-the-art methods in subtype identification accuracy and robustness.
Maintained superior performance under conditions with weak signal features and high noise levels.
Identified more survival-associated subtypes and clinically enriched features compared to competing methods.

Abstract

Abstract Motivation Cancer’s high heterogeneity necessitates precise molecular classification for improved clinical outcomes. However, current multi-omics clustering often struggles with molecular complexity. We propose Unified Latent and Similarity Learning (ULSL), a novel framework that simultaneously learns latent embeddings and similarity matrices through unified optimization. ULSL employs graph fusion for cross-omics structural consistency and latent representation learning to project data into low-dimensional spaces, effectively mitigating noise and high dimensionality. Results ULSL was evaluated on synthetic datasets and ten public cancer datasets from The Cancer Genome Atlas (TCGA). It consistently outperformed seven state-of-the-art methods in accuracy and robustness for subtype identification. On simulated datasets, ULSL maintained superior performance even with weak signal features and high noise levels. On TCGA datasets, ULSL not only identified survival-associated subtypes in a larger number of cancer types but also detected a greater number of clinically enriched features compared to competing approaches. Furthermore, the specific case study on AML demonstrated that ULSL aligns with the biological basis of the traditional FAB classification while offering distinct advantages in prognostic stratification. Availability and implementation The source code for ULSL is available at https://github.com/codelzy-01/ULSL-1.git

Read Full Paperexternally

Bookmark

View Full Paper

Cite This Study

Z et al. (Thu,) studied this question.

synapsesocial.com/papers/69e71423cb99343efc98d8c5 https://doi.org/https://doi.org/10.1093/bioadv/vbag110

Bookmark

View Full Paper