CMGL: Confidence-guided Multi-omics Graph Learning for Cancer Subtype Classification

TL;DR

CMGL introduces a confidence-guided multi-omics graph learning framework that enhances cancer subtype classification by estimating sample-specific modality reliability, leading to improved accuracy and better patient stratification.

Contribution

The paper presents a novel two-stage framework that estimates per-sample modality reliability and uses it to guide multi-omics data integration for cancer subtyping.

Findings

CMGL outperforms baseline methods with a 4.03% accuracy increase on four cancer tasks.

It accurately recovers PAM50 breast cancer subtypes.

The model trained on breast cancer data transfers effectively to kidney cancer for patient stratification.

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.