Iterative method improves integration of spatial multi-omics data
ILR-SMO: Iterative Latent Refinement for Robust Spatial Multi-Omics Integration
Machine Learning
Summary
Spatial multi-omics technologies measure different types of molecules in cells while keeping track of their location. Integrating these varied data types is tricky because they have different qualities and noise, which can cause errors. The authors propose ILR-SMO, a method that gradually combines these data step-by-step, checking and refining the combination each time. This approach makes the final result more stable and accurate, especially when the data are noisy or imbalanced. Tests on real datasets show their method works better than existing approaches.
What this means in practice
- •For bioinformatics engineers: Create more accurate combined views of spatial biological data by applying a stable iterative integration method that accounts for noisy and uneven data sources.
- •For pharmaceutical data teams: Improve analysis of complex tissue data to identify spatial cell domains more reliably during drug research by using an integration method resilient to data variability.
Authors
Anqi Yu, Xudong Xu, Jianzhi Lu, Yuqi Sun, Bingguo Chen, Chenxi Ma, Weimin Tan, Bo Yan
Abstract
Spatial multi-omics technologies jointly profile diverse molecular modalities with spatial context, providing a comprehensive view of cellular heterogeneity and tissue organization. To integrate spatial multi-omics data and identify spatial domains, a wide range of unsupervised methods has been proposed. However, recent approaches rely on single-step fusion, directly aggregating heterogeneous modalities into a shared representation, making embeddings sensitive to modality imbalance and measurement noise that are inherent to spatial multi-omics data, as well as unstable optimization under weak supervision. Here, we propose ILR-SMO (Iterative Latent Refinement for Spatial Multi-Omics), a unified self-supervised framework that reformulates multimodal integration as a stability-aware refinement process over a shared latent representation. Instead of single-step plain fusion, ILR-SMO progressively integrates modality-specific information through sequential updates, allowing information to be injected in a controlled manner. This incremental refinement enables later updates to correct or compensate for earlier deviations, resulting in more stable and robust representation learning under heterogeneous and noisy modalities. ILR-SMO further incorporates reliability-aware modality gating for adaptive modulation of modality contributions, and employs a joint objective to enforce spatial coherence while preventing representation collapse. Extensive experiments on five spatial multi-omics benchmarks demonstrate that ILR-SMO consistently outperforms seven state-of-the-art methods and exhibits strong robustness across diverse settings.