Installation

This vignette describes how to install NEBULA and its dependencies on a local machine.

System requirements

NEBULA relies on a C++ backend compiled through Rcpp. Before installing, ensure the following are available on your system:

R >= 4.1.0
A working C++17 compiler:
- Linux: gcc / g++ (install via apt install g++ or equivalent)
- macOS: Xcode Command Line Tools (xcode-select --install)
- Windows: Rtools
PLINK (version 1.9 or 2.0) must be available on the system PATH if you are working with real genotype data. PLINK is not required for running the synthetic data workflow.

Installing NEBULA from GitLab

NEBULA is hosted on GitLab at https://gitlab.com/sysbiobig/nebula.

Install with vignettes

The recommended installation command builds the vignettes locally so that they can be accessed offline via browseVignettes("nebula"):

install.packages("devtools")

devtools::install_gitlab(
  "sysbiobig/nebula",
  build_opts = c("--no-resave-data", "--no-manual"),
  build_vignettes = TRUE
)

Install without vignettes

If you do not need local copies of the vignettes (e.g. in a scripted pipeline or a container build), use the lighter installation:

devtools::install_gitlab("sysbiobig/nebula")

R package dependencies

NEBULA will install its R dependencies automatically. The main ones are:

Package	Role
`Rcpp`	C++ interface
`BH`	Boost C++ headers
`data.table`	Fast data manipulation
`readr`	File I/O
`dplyr`	Data wrangling
`tidyr`	Data reshaping
`foreach`	Parallel iteration
`doParallel`	Parallel back-end
`genio`	PLINK binary file I/O
`biomaRt`	Bioconductor annotation (used in correction)
`fuzzyjoin`	Approximate join for gene annotation

biomaRt is a Bioconductor package. If it is not installed automatically, run:

install.packages("BiocManager")
BiocManager::install("biomaRt")

Verifying the installation

After installation, load the package and check the version:

library(nebula)
packageVersion("nebula")

Browse the available vignettes:

browseVignettes("nebula")

List the exported functions: