Subtype identification from heterogeneous TCGA datasets on a genomic scale by multi-view clustering with enhanced consensus

Abstract

BackgroundThe Cancer Genome Atlas (TCGA) has collected transcriptome, genome and epigenome information for over 20 cancers from thousands of patients. The availability of these diverse data types makes it necessary to combine these data to capture the heterogeneity of biological processes and phenotypes and further identify homogeneous subtypes for cancers such as breast cancer. Many multi-view clustering approaches are proposed to discover clusters across different data types. The problem is challenging when different data types show poor agreement of clustering structure.ResultsIn this work, we first propose a multi-view clustering approach with consensus (CMC), which tries to find consensus kernels among views by using Hilbert Schmidt Independence Criterion. To tackle the problem when poor agreement among views exists, we further propose a multi-view clustering approach with enhanced consensus (ECMC) to solve this problem by decomposing the kernel information in each view into a consensus part and a disagreement part. The consensus parts for different views are supposed to be similar, and the disagreement parts should be independent with the consensus parts. Both the CMC and ECMC models can be solved by alternative updating with semi-definite programming. Our experiments on both simulation datasets and real-world benchmark datasets show that ECMC model could achieve higher clustering accuracies than other state-of-art multi-view clustering approaches. We also apply the ECMC model to integrate mRNA expression, DNA methylation and microRNA (miRNA) expression data for five cancer data sets, and the survival analysis show that our ECMC model outperforms other methods when identifying cancer subtypes. By Fisher’s combination test method, we found that three computed subtypes roughly correspond to three known breast cancer subtypes including luminal B, HER2 and basal-like subtypes.ConclusionIntegrating heterogeneous TCGA datasets by our proposed multi-view clustering approach ECMC could effectively identify cancer subtypes.

Convex Multi-View Clustering Via Robust Low Rank Approximation With Application to Multi-Omic Data

“…Moreover, Figure 4 also shows the misclassified samples when clustering on the CGRMSL subspace, and misclassified samples by k-means on the original space before dimensionality reduction. Different cancer subtypes are expected to have significantly different survival times [4]. Here we apply our model to identify potential cancer subtypes by performing a survival analysis on the obtained clusters.…”

“…R ECENT advances in high throughput sequencing technologies have made available large amounts of biomedical data consisting of measurements of genomic features across multiple omic scales forming multi-omic datasets when combined. Multi-omic data have been recently used to efficiently visualize and cluster cancer subtypes [4]. Clustering for biomedical data is a useful pattern discovery technique, which is the initial step taken in data exploration.…”

A Hierarchical Clustering Algorithm Based on Silhouette Index for Cancer Subtype Discovery from Omics Data

“…The paper [6] introduced a multi-view clustering approach (CMC) and its enhanced version (ECMC) for subtype identification from heterogeneous cancer datasets. The authors employed both Spectral Clustering and K-Means together with Silhouette Index to identify novel cancer subtypes.…”

A bioinformatics potpourri

“… Case study of network bi-stability and positive/negative feedback loops in TGF-β1 activation [ 18 ]. Cancer and disease informatics Multi-view clustering method with enhanced consensus; breast cancer sub-typing and survival analysis [ 19 ]. Network hubs as prognostic signatures in ovarian cancer, breast cancer and glioblastoma multiforme selected by Cox regression for correlating DNA methylation levels with outcome [ 20 ].…”