lncCHDNet - Long Non-Coding RNAs (lncRNAs) Congenital Heart Disease Network analysis

Author: J. Penaloza and Peter White
Date: September 18, 2024

Department: The Office of Data Sciences
Organization: Abigail Wexner Research Institute, Nationwide Children's Hospital
Address: 575 Children's Crossroad, Columbus, OH 43215 USA

This repository contains R notebooks for the analysis of long non-coding RNAs (lncRNAs) associated with copy number variants (CNVs) in congenital heart disease (CHD) patients from the CCVM cohort. The analysis includes transcript quantification, network construction using WGCNA, and functional enrichment analysis of lncRNAs involved in CHD. This code is also stably archived at

Project Overview

In this project, we aim to investigate the role of lncRNAs impacted by CNVs in CHD. Using the WGCNA method, we identified key gene modules and candidate lncRNAs associated with heart development. This repository contains R Markdown notebooks and rendered HTML files that guide the data processing, analysis, and visualization steps.

Repository Contents

1. Penaloza_CCVM_Notebook_1A.Rmd: R-notebook preprocessing of datasets and pinpointing lncRNAs within clinically validated CNVs.

2. Penaloza_CCVM_Notebook_1B.Rmd: R-notebook for co-expression networks, identification and functional analysis of lncRNA-CNVs.

   • InputData: This folder contains the CCVM data file from which all analysis is dependent.
     • 02032023_CCVM_abnormalEcho.Rds is a binary R file required for the script to run. 02032023_CCVM_abnormalEcho.xlsx is the data in Microsoft Excel format.
   • PublicData: This zip folder contains all of the publically available data that is required for the analysis.
     • Available for dowload on zenodo:
     • This includes:
       • lncExpDB_E-MTABGeneTPM.tsv: Normalized TPM values for the developmental time series data
       • E-MTAB-6814.sdrf.txt: Metadata for the developmental time series
       • LncBookv1.9_GENCODEv33_GRCh38.gtf.gz: GTF file used for RNA-Seq transcript quantification
       • UCSC_hg19ToHg38.over.chain: Chain file to liftover genomic coordinates from GRCh37 to GRCh38
       • LncBook_id_conversion.csv: LncBook provided conversions of LncBook accessions
       • RNACentral_lncbook.tsv: RNACentral to LncBook accessions
       • RNACentral_ensembl.tsv: RNACenral to Ensembl accessions
       • chdgene_table.csv: List of 142 known CHD genes from CHDgene website
       • Cotney_CircRes_316709_online_table_v.xlsx: VanOudenhove et al., 2020. Table V
   • Results: This folder will contain output files such as gene module lists, hub gene lists, and GO enrichment analysis results after running the scripts.
   • Figures: This folder will contain plots and visualizations generated during the analysis. All are found in our manuscript.

Dependencies

The following R packages are required to run the analysis:

Data Manipulation and Tidy Data

• tidyr
• dplyr
• stringr
• purrr
• tibble
• readr
• data.table
• reshape2
• foreach
• doParallel

Visualization and Plotting

• ggplot2
• grid
• gridExtra
• ggdendro
• ggrepel
• Cairo
• viridis
• gplots
• ggpubr
• circlize
• eulerr

Genomic Data and Network Analysis

• GenomicRanges
• biomaRt
• rtracklayer
• WGCNA
• RCy3
• igraph
• clusterProfiler
• GOSemSim
• org.Hs.eg.db
• gprofiler2

File Handling and Export

• readxl
• writexl
• R.utils
• htmlwidgets
• webshot2
• chromote

Reporting and Formatting

• knitr
• kableExtra
• rmdformats
• skimr
• here

Packages available on CRAN are installed using install.packages(), and those available on Bioconductor are installed using BiocManager::install().

CRAN Packages:

# Install CRAN packages
install.packages(c(
  "tidyr", "dplyr", "stringr", "purrr", "tibble", "readr", "data.table",
  "reshape2", "foreach", "doParallel", "ggplot2", "grid", "gridExtra",
  "ggdendro", "ggrepel", "Cairo", "viridis", "gplots", "ggpubr", "circlize",
  "eulerr", "readxl", "writexl", "R.utils", "htmlwidgets", "webshot2",
  "chromote", "knitr", "kableExtra", "rmdformats", "skimr", "here"
))

Bioconductor Packages:

# Install Bioconductor packages
if (!requireNamespace("BiocManager", quietly = TRUE))
  install.packages("BiocManager")

BiocManager::install(c(
  "GenomicRanges", "biomaRt", "rtracklayer", "WGCNA", "RCy3", "clusterProfiler",
  "GOSemSim", "org.Hs.eg.db", "gprofiler2"
))

Explanation:

• CRAN Packages: These packages are installed using install.packages(). Simply run the first command in R to install all the CRAN packages in one go.
• Bioconductor Packages: Bioconductor packages are installed using BiocManager::install(). Make sure the BiocManager package is installed first, and then install the Bioconductor-specific packages listed.

This should install all the necessary dependencies for the CNV-lncRNA analysis project.

Setup

Ensure that the data files required for analysis (e.g., TPM matrices, metadata) are available in the appropriate directories specified in the scripts.

Usage

1. Penaloza_CCVM_Notebook_1A.Rmd: Preprocessing and Data Normalization

   • Run this notebook to preprocess RNA-Seq data, normalize the TPM values, and threshold the data for WGCNA analysis.

   Command (in RStudio or terminal):

   rmarkdown::render("Penaloza_CCVM_Notebook_1A.Rmd")

2. Penaloza_CCVM_Notebook_1B.Rmd: WGCNA and Hub Gene Identification

   • Run this notebook to perform WGCNA, identify heart-specific gene modules, and find hub genes.

   Command (in RStudio or terminal):

   rmarkdown::render("Penaloza_CCVM_Notebook_1B.Rmd")

Output

• Gene Modules: Lists of genes associated with each WGCNA module, saved in /Results.
• Hub Genes: Identified hub genes within key heart-related modules.
• GO Enrichment Analysis: Functional enrichment results for modules, including GO biological processes, molecular functions, and cellular components.
• Figures: Key plots such as gene module dendrograms, eigengene expression patterns, and co-expression networks (saved in /Figures).
• Results: The results of the filtering and liftover process from Notebook 1A, Penaloza_lncRNA_CCVM_hg38_liftover.rds and the intersection of this data with expressed lncRNAs, Penaloza_expressed_lncRNA_CCVM_hg38_liftover.rds. The results of processing the LncBook GTF (NCH_LncBookv1.9_GENCODEv33_GRCh38.gtf.gz and NCH_LncBookv1.9_GENCODEv33_GRCh38.gff.gz). Supplemental Data Tables and results tables (saved in /Results).

Network Visualization

The function create_and_view_network() within the notebook Penaloza_CCVM_Notebook_1B.Rmd allows for the creation and visualization of gene co-expression networks. The network can be viewed using Cytoscape, and the function saves the network as a GraphML file for further analysis.