Single-Cell Co-expression Analysis Reveals Distinct Functional Modules, Co-regulation Mechanisms and Clinical Outcomes

Abstract

Co-expression analysis has been employed to predict gene function, identify functional modules, and determine tumor subtypes. Previous co-expression analysis was mainly conducted at bulk tissue level. It is unclear whether co-expression analysis at the single-cell level will provide novel insights into transcriptional regulation. Here we developed a computational approach to compare glioblastoma expression profiles at the single-cell level with those obtained from bulk tumors. We found that the co-expressed genes observed in single cells and bulk tumors have little overlap and show distinct characteristics. The co-expressed genes identified in bulk tumors tend to have similar biological functions, and are enriched for intrachromosomal interactions with synchronized promoter activity. In contrast, single-cell co-expressed genes are enriched for known protein-protein interactions, and are regulated through interchromosomal interactions. Moreover, gene members of some protein complexes are co-expressed only at the bulk level, while those of other complexes are co-expressed at both single-cell and bulk levels. Finally, we identified a set of co-expressed genes that can predict the survival of glioblastoma patients. Our study highlights that comparative analyses of single-cell and bulk gene expression profiles enable us to identify functional modules that are regulated at different levels and hold great translational potential.

Fig 1. Distinct sets of co-expressed genes were identified for single cells and bulk tissues.

(A) Examples of single-cell specific, bulk specific and shared co-expressed gene pairs. (B) The mapping of the top 1,000 positively correlated genes in single cells (or bulk tissues) to their correlation in bulk tissues (or single cells). Each line represent a gene pair. The cells (or tissues) were randomly partitioned into two sub-groups. (C) The mapping of the top 1,000 positively correlated genes in one sub-group to the other sub-group. (D) Clustering of glioblastomas based on gene expression profiles. Bulk samples from TCGA (TCGA-GBM), single-cell-derived average gene expression (Average-MGH) and 5 genuine bulk samples (Bulk-MGH) for single-cell sequencing.

Fig 2. Members in protein complexes are predominately connected by one type of co-expressions.

(A) The fraction of co-expressed genes whose protein products interact with each other. (B) The same fraction in function of correlation coefficients. (C) Examples of protein complexes. Two gene members in a complex were connected if they are co-expressed. The color denotes the types of co-expressions: single-cell specific (orange), bulk specific (cyan), and shared (green).

Fig 3. Different types of co-expressions are associated with distinct biological functions.

(A) The fraction of co-expressed genes that have the similar biological functions. (B) The same fraction in function of correlation coefficients. (C) Gene function network of top 1,000 co-expressions. Genes with the same functions are placed around circles. Two genes are connected in lines if they have single-cell specific (orange), bulk specific (cyan), or shared (green) co-expression.

Fig 4. Distinct regulatory mechanisms are associated with co-expressions in single cells and bulk tissues.

(A) Two models for co-regulation. Two genes which were detected to have synchronized promoters at the bulk level may not be simultaneously regulated at the single-cell level. Two genes interacting with each other in 3-D chromatin may be co-regulated by the same enhancer. (B) Examples of DHS correlations in three types of co-expressed gene pairs. The figure showed six cell lines as examples. The correlation coefficients (R) were calculated based on 125 cell lines. (C) The distribution of the correlation coefficients of DHS signals across 125 cell types between co-expressed genes. (D) Fraction of co-expressed genes that have genomic interactions.

Fig 5. Interchromosomal interactions are prevalent in single-cell co-expressed genes.

(A) The fraction of co-expressed gene pairs located in the same chromosomes. (B) Map of the co-expressed gene pairs. Each line represent a pair of co-expressed genes. The color of the lines indicates whether they are in the same chromosomes: purple (intrachromosomal), and gray (interchromosomal). (C) The fraction of co-expressed gene pairs located in the same topological domains. (D) The fraction of intrachromosomal and interchromosomal co-expression genes with chromosomal interactions.

Fig 6. A set of co-expressed genes can serve as a prognosis signature for glioblastomas.

(A) Flow chart of the selection of co-expressed genes for prognosis analysis. (B) Six genes were selected to classify the patients. Four genes are shared co-expressed (green), and two genes are single-cell specific co-expressed (orange). (C) Kaplan-Meier survival curves in two groups of 120 sequencing samples. Log-rank test was used. (D) Survival curves for an independent validation set.