CMGL: Confidence-guided Multi-omics Graph Learning for Cancer Subtype Classification
Boyang Fan, Hengchuang Yin, Siyu Yi, Yifan Wang, Zhicheng Li + 3 more
TLDR
CMGL is a two-stage framework that uses evidential deep learning to estimate per-sample modality reliability, improving cancer subtype classification.
Key contributions
- Introduces CMGL, a two-stage framework for robust multi-omics cancer subtyping.
- Estimates per-sample modality reliability via evidential deep learning to guide data fusion.
- Outperforms baselines by 4.03% in accuracy on single-cancer subtype tasks.
- Demonstrates recovery of known subtypes and successful transfer learning to new cancer types.
Why it matters
This paper addresses a key challenge in multi-omics cancer subtyping: varying data quality and noise. By independently estimating modality reliability, CMGL prevents low-quality data from distorting patient similarity graphs. Its improved accuracy and transferability offer a more robust approach for precision oncology.
Original Abstract
Motivation: Multi-omics integration can improve cancer subtyping, but modality informativeness and noise vary across cancer types and patients. Existing graph-based methods optimize modality weights jointly with the classification objective and therefore lack independent reliability estimates, so low-quality omics distort patient similarity graphs and amplify noise through message passing. Results: We propose CMGL, a two-stage framework that estimates per-sample modality reliability through evidential deep learning and uses the frozen confidence scores to guide cross-omics fusion and graph construction. On four MLOmics cancer-subtype tasks and the 32-class pan-cancer task, CMGL consistently improves over the strongest baseline, surpassing it by 4.03% in average accuracy on the four single-cancer tasks. Its representations recover the PAM50 intrinsic subtypes of breast invasive carcinoma (BRCA), and the BRCA-trained model transfers without fine-tuning to kidney renal clear cell carcinoma (KIRC), stratifying patients into prognostically distinct groups.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.