Bayesian network-driven clustering analysis with feature selection for high-dimensional multi-modal molecular data

Yize Zhao,Ronglai Shen,Margaret Hannum,Jasme Lee,Changgee Chang

doi:10.1038/s41598-021-84514-0

Abstract

Multi-modal molecular profiling data in bulk tumors or single cells are accumulating at a fast pace. There is a great need for developing statistical and computational methods to reveal molecular structures in complex data types toward biological discoveries. Here, we introduce Nebula, a novel Bayesian integrative clustering analysis for high dimensional multi-modal molecular data to identify directly interpretable clusters and associated biomarkers in a unified and biologically plausible framework. To facilitate computational efficiency, a variational Bayes approach is developed to approximate the joint posterior distribution to achieve model inference in high-dimensional settings. We describe a pan-cancer data analysis of genomic, epigenomic, and transcriptomic alterations in close to 9000 tumor samples across canonical oncogenic signaling pathways, immune and stemness phenotype, with comparisons to state-of-the-art clustering methods. We demonstrate that Nebula has the unique advantage of revealing patterns on the basis of shared pathway alterations, offering biological and clinical insights beyond tumor type and histology in the pan-cancer analysis setting. We also illustrate the utility of Nebula in single cell data for immune cell decomposition in peripheral blood samples.

Highlights

Multi-modal molecular profiling data in bulk tumors or single cells are accumulating at a fast pace
The joint distribution of (γ, θ 1, θ0) discriminates the subjects by clusters under a mixture of null and alternative distributions, and it follows a distribution G, which samples from a Dirichlet process mixture (DPM) with base measure G0 and concentration parameter α0
The discrete nature of DPM realizes a distribution-based grouping effect, where each group is defined only under its specific active biomarker set across modalities denoted as (X11, . . . , XM1 )

Summary

Introduction

Multi-modal molecular profiling data in bulk tumors or single cells are accumulating at a fast pace. With recent advances in multi-modality profiling of single c ells[2,3,12,13], integrative clustering algorithms with the incorporation of network priors will allow more in-depth understanding of cell heterogeneities and interactions. Scalable implementation is crucial for clustering analysis of modern multi-model data sets across tens of thousands of patient samples or single cells. To address these challenges, we aim to develop a high-dimensional clustering method that incorporates biological network information across different data modalities. We present Nebula (Network-based multi-modal clustering analysis), a novel Bayesian network-based clustering analysis for multi-modal integration and clustering with feature selection, and compare its performance in pan-cancer tumor profiling and single cell transcriptome sequencing data to state-of-the-art clustering algorithms including iCluster[14] and Suerat[8]

Objectives

Methods

Results

Conclusion