Reconstruction of Cell-type-Specific Interactomes at Single-Cell Resolution.

Shahin Mohammadi,Jose Davila-Velderrain,Manolis Kellis

doi:10.1016/j.cels.2019.10.007

Abstract

The human interactome is instrumental in the systems-level study of the cell and the contextualization of disease-associated gene perturbations. However, reference organismal interactomes do not capture the cell-type-specific context in which proteins and modules preferentially act. Here, we introduce SCINET, a computational framework that reconstructs an ensemble of cell-type-specific interactomes by integrating a global, context-independent reference interactome with a single-cell gene-expression profile. SCINET addresses technical challenges of single-cell data by robustly imputing, transforming, and normalizing the initially noisy and sparse expression of data. Inferred cell-level gene interaction probabilities and group-level interaction strengths define cell-type-specific interactomes. We use SCINET to reconstruct and analyze interactomes of the major human brain and immune cell types, revealing specificity and modularity of perturbations associated with neurodegenerative, neuropsychiatric, and autoimmune disorders. We report cell-type interactomes for brain and immune cell types, together with the SCINET package.

Highlights

Proteins participate in cross-talking pathways and overlapping functional modules that collectively mediate cell behavior
Methodological Overview The core SCINET framework (Figure 1) is based on the following methodological developments: (1) a decomposition method to interpolate values for missing observations in the scRNA-seq profile, (2) a parametric approach to project heterogeneous gene expression distributions into a compatible subspace (Figure 1B), (3) a statistical framework to measure the likelihood of gene interactions within each cell, and (4) a subsampling approach to aggregate interaction likelihoods of individual cells, reduce noise, and estimate the underlying distribution and variability of interaction strengths within each cell-type population (Figure 1C)
A single-cell expression matrix is iteratively decomposed into lower dimensional matrices at different levels of granularity defined by a number of low-rank factors

Summary

Introduction

Proteins participate in cross-talking pathways and overlapping functional modules that collectively mediate cell behavior. One approach to incorporate context into a global interactome network is by considering the transcriptional dynamics of the network’s components, effectively integrating a static snapshot of the space of all potential interactions with context-specific gene expression. This approach has been previously applied to construct tissue-specific networks, taking advantage of bulk gene expression measurements (Mohammadi and Grama, 2016; Magger et al, 2012; Bossi and Lehner, 2009). The increasing availability of these data provides a unique opportunity to study the context specificity of molecular interactions at single-cell resolution. The development of efficient and robust techniques is required to transition to a single-cell network biology

Methods

Results

Discussion

Conclusion