TL;DR
Researchers have developed ExpressHeart, a user-friendly web portal that integrates and visualizes single-cell RNA sequencing (scRNA-seq) data from non-cardiomyocyte (nonCM) cells in the heart across humans, mice, and zebrafish. This tool addresses the challenge of accessing and analyzing the growing body of nonCM scRNA-seq data, enabling deeper insights into heart disease mechanisms and gene function across species.
What Problem Does This Paper Address?
The heart is more than just muscle cells (cardiomyocytes); it's a complex organ with various non-cardiomyocyte (nonCM) cells like fibroblasts, endothelial cells, and immune cells that are crucial for structural support, mechanical function, and overall cardiac health. Understanding the molecular roles of these nonCMs is vital for studying heart diseases. While single-cell RNA sequencing (scRNA-seq) has revolutionized our ability to study these cells at an unprecedented resolution, the sheer volume of data generated from different labs, species, and experimental conditions makes it difficult for researchers to access, integrate, and analyze this information effectively. There was a clear need for a centralized platform to visualize and compare these diverse nonCM datasets.
Methods
ExpressHeart is built as a web application using R Shiny, providing an interactive interface for users. The core functionality involves:
- Data Integration: It consolidates scRNA-seq and single-nucleus RNA-seq (snRNA-seq) data from five distinct studies.
- Dimensionality Reduction and Visualization: Standard techniques like Uniform Manifold Approximation and Projection (UMAP) and t-distributed Stochastic Neighbor Embedding (t-SNE) are used to visualize cell populations. Users can choose between these methods and download the resulting plots.
- Gene Expression Visualization: Users can explore gene expression patterns through:
- Violin Plots: To compare gene expression ranges and densities across different cell types/subtypes.
- Feature Plots: To visualize gene expression distribution on UMAP/t-SNE plots, showing spatial patterns.
- Heatmaps: To display the expression levels of multiple genes across cells, grouped by cell type/subtype.
- Differential Gene Expression (DEG) Analysis: The portal identifies DEGs within and between cell types/subtypes for each dataset.
- Cross-Dataset and Cross-Species Comparison: This is a key feature, allowing users to:
- Query Genes: Check if a specific gene is differentially expressed across datasets.
- Identify High-Confidence DEGs: Find genes consistently identified as differentially expressed across multiple datasets within the same species, aiding in robust cell type identification.
- Compare Homolog Gene Expression: Visualize and compare the expression patterns of homologous genes across different species, revealing conserved and divergent functions.
Datasets
ExpressHeart integrates data from five scRNA-seq/snRNA-seq datasets:
- Human-Hocker-2021: Healthy adult human hearts (snRNA-seq).
- Mouse-Wang-2021: Normal adult mouse hearts (scRNA-seq).
- Mouse-Farbehi-2019: Control and myocardial infarction (MI) injured mouse hearts (scRNA-seq).
- Mouse-McLellan-2020: Untreated, sham-, and Angiotensin II-treated adult mouse hearts (scRNA-seq).
- Zebrafish-Ma-2021: Uninjured (control) and injured zebrafish hearts (scRNA-seq).
These datasets cover different species (human, mouse, zebrafish) and conditions (healthy, injury, disease models), providing a rich resource for comparative analysis.
Key Insights and Ideas
- Bridging the Data Silo: The most significant contribution is creating a unified platform for disparate nonCM scRNA-seq data, making complex analyses accessible to a broader research community.
- Unveiling Species-Specific Gene Roles: The cross-species comparison highlights how homologous genes can have different expression patterns and potentially different functional roles in different species. For example, the gene Mmp14 (mouse) and mmp14b (zebrafish) show distinct expression in macrophages between mice and zebrafish, suggesting a role in immune processes in zebrafish that is less pronounced in mice.
- Enhancing Cell Type Annotation: Identifying "high-confidence DEGs" shared across multiple datasets within a species can improve the accuracy and reproducibility of cell type and subtype identification in new scRNA-seq studies.
- Facilitating Subtype Matching: The ability to compare expression profiles across datasets allows researchers to match similarly labeled subtypes from different studies, even if they have custom labels, leading to a better understanding of functional states.
Key Results
- ExpressHeart successfully integrates and visualizes scRNA-seq data from five diverse datasets across three species.
- The platform allows users to explore gene expression patterns, identify DEGs, and perform cross-dataset and cross-species comparisons.
- The tool demonstrated its utility by identifying shared DEGs between mouse datasets and highlighting species-specific expression patterns of homologous genes. For instance, it showed that Mmp14 in mice is primarily expressed in fibroblasts, while its zebrafish homolog mmp14b is expressed in both fibroblasts and macrophages.
Limitations and Future Work
- Scope: Currently, ExpressHeart focuses solely on transcriptomic data (scRNA-seq/snRNA-seq). The authors acknowledge that incorporating other omics data, such as single-nucleus ATAC-seq (snATAC-seq) for open chromatin status, DNA methylation, or chromatin interactions, would provide a more comprehensive view of nonCM molecular features.
- Data Expansion: While the current version includes five datasets, the authors suggest that future work could involve incorporating more multi-omics single-cell assays to capture a broader spectrum of cellular information.
- User-Driven Analysis: The platform provides pre-computed analyses and visualization tools. Future iterations could potentially offer more advanced user-driven analytical capabilities.
Summary updated