intCC: An efficient weighted integrative consensus clustering of multimodal data.

Area of Science:

• Computational biology
• Bioinformatics
• Machine learning for omics data analysis

Background:

• High-throughput multiomics data offers insights into complex diseases like cancer.
• Identifying distinct cancer subtypes is crucial for targeted therapies and improved patient outcomes.
• Integrative clustering is a key unsupervised learning approach for subtype discovery.

Purpose of the Study:

• To develop an efficient weighted integrative clustering method named intCC.
• To enhance the accuracy of subtype discovery from multiomics data.
• To provide a robust computational tool for cancer research.

Main Methods:

• Combining ensemble methods, consensus clustering, and kernel learning for integrative clustering.
• Developing the intCC algorithm for efficient and accurate cluster analysis.
• Utilizing extensive simulation studies and real-world cancer datasets (TCGA).

Main Results:

• intCC effectively uncovers latent cluster structures in multiomics data.
• The method demonstrates high accuracy in identifying potential cancer subtypes.
• A case study on TCGA pan-cancer datasets validates the performance of intCC.

Conclusions:

• intCC is a powerful and efficient tool for integrative clustering and cancer subtype discovery.
• The proposed method aids in understanding the complexity of human diseases through multiomics data analysis.
• An R package for intCC is publicly available for broader research application.