Integrative cancer patient stratification via subspace merging

Abstract

Motivation: Technologies that generate high-throughput omics data are flourishing, creating enormous, publicly available repositories of multi-omics data. As many data repositories continue to grow, there is an urgent need for computational methods that can leverage these data to create comprehensive clusters of patients with a given disease. Results: Our proposed approach creates a patient-to-patient similarity graph for each data type as an intermediate representation of each omics data type and merges the graphs through subspace analysis on a Grassmann manifold. We hypothesize that this approach generates more informative clusters by preserving the complementary information from each level of omics data. We applied our approach to The Cancer Genome Atlas (TCGA) breast cancer dataset and show that by integrating gene expression, microRNA and DNA methylation data, our proposed method can produce clinically useful subtypes of breast cancer. We then investigate the molecular characteristics underlying these subtypes. We discover a highly expressed cluster of genes on chromosome 19p13 that strongly correlates with survival in TCGA breast cancer patients and validate these results in three additional breast cancer datasets. We also compare our approach with previous integrative clustering approaches and obtain comparable or superior results.

Clustering of cancer data based on Stiefel manifold for multiple views

“…In particular, AASC and MVSC method can effectively identify different clusters. For the methods based on manifold, we chose MOCMO [ 17 ] and Grassmann manifold clustering method [ 18 ]. For the state-of-the-art methods, we chose MvNE algorithm [ 19 ].…”

“…In this paper, “Heat Kernel” is used to measure the similarity between samples [ 18 ]. The basic form is a Gaussian function with “t”.…”

“…Essentially, we assume that local similarity is more reliable than remote similarity. This is a modest assumption, and it is widely used by other manifold learning algorithms [ 18 ].…”

Clustering of Cancer Data Based on Stiefel Manifold for Multiple Views

“…Essentially, we assume that local similarity is more reliable than remote similarity. This is a modest assumption, and it is widely used by other manifold learning algorithms (Ding et al, 2018).…”

“…If the current U does not satisfy inequation (11), then a parameter τ is required to adjust the step size:…”