SCovid v2.0: a comprehensive resource to decipher the molecular characteristics across tissues in COVID-19 and other human coronaviruses

Abstract

SCovid v2.0 (http://bio-annotation.cn/scovid or http://bio-computing.hrbmu.edu.cn/scovid/) is an updated database designed to assist researchers in uncovering the molecular characteristics of coronavirus disease 2019 (COVID-19) across various tissues through transcriptome sequencing. Compared with its predecessor, SCovid v2.0 is enhanced with comprehensive data, practical functionalities, and a reconstructed pipeline. The current release includes (i) 3,544,360 cells from 45 single-cell RNA-seq (scRNA-seq) data sets encompassing 789 samples from 15 tissues; (ii) the addition of 62 COVID-19 bulk RNA-seq data comprising 1,688 samples from 12 tissues; (iii) incorporation of seven bulk RNA-seq data sets related to other human coronaviruses, such as HCoV-229E, HCoV-OC43, and MERS-CoV for a thorough comparative analysis of pan-coronavirus mechanisms in COVID-19; and (iv) systematic comparisons between the data sets conducted using standardized procedures. Furthermore, we have developed an advanced search engine and upgraded web interface to browse, search, visualize, and download detailed information. Overall, SCovid v2.0 is a valuable resource for exploring molecular characteristics of COVID-19 across different tissues.

Importance: This manuscript provides a comprehensive analysis of the molecular characteristics of COVID-19 through cross-tissue transcriptome analysis, contributing to the understanding of COVID-19 by clinicians and scientists. Considering the cyclical nature of coronavirus outbreaks, this updated database adds transcriptome data on other human coronaviruses, contributing to potential and existing mechanisms of other human coronaviruses.

Keywords: COVID-19; bulk RNA-seq; molecular characteristics; other human coronaviruses; single-cell RNA-seq.

Fig 1

Statistics of projects and samples among organisms. Data set statistics for different tissues in (A) COVID-19|single-cell RNA-seq; (B) COVID-19|bulk RNA-seq; and (C) other human coronaviruses|bulk RNA-seq. (D) The experimental and control group statistics for single-cell RNA-seq and bulk RNA-seq data.

Fig 2

Overview of data collection, analysis workflow, and database function. (A) The source of SCovid data. (B) The workflow of SCovid. Here is an example of the single-cell RNA-seq workflow. (C) The functional modules of SCovid are a user-friendly web interface for users to browse data set information, search genes in single or multiple data sets, and acquire data sets.

Fig 3

Database functions on the Browse page of SCovid v2.0. An example is the overview of GSE186267 (Eddins et al. [endotracheal aspirates]). (A) Two-dimensional tSNE plot. The colors of points represent the cell- (left) or sub-type (right) to which cells belong. (B) Cell proportion plot that displays the proportion of each cell- (left) or sub-type (right) per sample in the selected data set. (C) The volcano plot shows the statistically significant DEGs between COVID-19 and control. (D) KEGG classifications of downregulated genes (e.g., epithelial). The vertical axis shows the names of clusters of KEGG terms, and the horizontal axis displays the -Log₁₀ (P-value). A P-value < 0.05 was used as a threshold to select significant KEGG terms. (E) Expression from each group and tSNE projection (e.g., ACTB).

Fig 4

Downstream analysis on the Browse page of SCovid v2.0. (A) The heatmap shows the expression profile of high-variance genes in different cell- (left) or sub-types (right). (B) The cell differentiation plots show cell differentiation trajectories in cell-types (e.g., epithelial). (C) The two types of cell communication plots explain cell–cell interaction and interaction strength between different cell- (left) or sub-types (right).