4  Installing programmes

Note⏳ Time
  • Teaching: 10 min
  • Exercises: 5 min
Note🤔 Questions
  • How do software packages get installed on a Unix system?
  • How do I check whether a programme is already available?
  • How do I install something from the command line?
Note🎯 Objectives
  • Use which to check whether a programme is on your PATH
  • Check a programme’s version with --version
  • Install a package using the system package manager
  • Install NCBI Entrez Direct using a shell installer script

4.1 Is it already there? (Checking availability)

Before installing anything, it is worth checking whether the programme is already available. The command which tells you where on the system a given programme lives:

which curl

If the programme is installed, which prints its path, e.g.:

/usr/bin/curl

If it is not found, which prints nothing (or an error). You can also check the version of an installed programme — most accept either --version or -v:

curl --version
Tip

Running which and --version before installing anything is a good habit. Many tools are already present on your system, and reinstalling them unnecessarily can cause conflicts.

4.2 How installation works on Unix

When you type a command, the shell searches a list of directories called the PATH to find the corresponding programme. If it finds it, it runs it. If not, you get command not found.

Installing a programme means placing it (or a link to it) somewhere on your PATH. There are several common ways to do this:

System package managers

The most common approach is to use the system package manager, which handles downloading, installing, and keeping track of software:

System Package manager Example
Ubuntu / Debian apt sudo apt install tree
Fedora / RHEL dnf sudo dnf install tree
macOS (Homebrew) brew brew install tree
Windows (MobaXterm) apt apt install tree

The word sudo (superuser do) runs the command with administrator privileges, required when installing software system-wide. MobaXterm users do not need sudo — it runs in a user-level environment where apt works directly.

Shell installer scripts

Some tools are distributed as installer scripts that you download and run directly in the shell. This is the approach used by NCBI Entrez Direct, Conda, and many others. The general pattern is:

curl -fsSL https://example.com/install.sh | sh

This downloads an install script with curl and pipes it directly into the shell to execute. We will use exactly this pattern to install Entrez Direct below.

Warning

Piping a script from the internet directly into your shell is convenient, but you are trusting the source. Only do this with scripts from well-known, reputable organisations (NCBI, Conda, Homebrew, etc.).

4.3 Installing NCBI Entrez Direct

NCBI Entrez Direct (EDirect) is a set of command-line tools for querying NCBI databases, including downloading sequences, searching PubMed, and retrieving metadata. We will use it in the next chapter to download a genome directly from the command line.

Step 1: Check if it is already installed

which efetch

If you get a path back (e.g. ~/edirect/efetch), it is already installed — skip to Step 3.

Step 2: Run the installer

sh -c "$(curl -fsSL https://ftp.ncbi.nlm.nih.gov/entrez/entrezdirect/install-edirect.sh)"

Breaking this down:

Part Meaning
curl -fsSL <url> Download the installer script silently
-f Fail silently on server errors
-s Silent mode (no progress bar)
-S Show errors even in silent mode
-L Follow redirects
sh -c "$(…)" Execute the downloaded script in the current shell

When prompted, allow the installer to add edirect to your PATH.

Step 3: Reload your PATH

After installation, close and reopen your terminal, or run:

export PATH="${HOME}/edirect:${PATH}"

This temporarily adds the edirect directory to your PATH for the current session. The installer should have added it permanently to your shell configuration file (.bashrc or .zshrc) for future sessions.

Step 4: Verify the installation

which efetch
efetch --help | head -n 5

You should see the path to efetch and the beginning of its help text.

Note

Windows / MobaXterm users: the installer may not complete successfully. If which efetch returns nothing after running the installer, use the curl backup method in the next chapter, which will work on all systems.

Caution✏️ Exercise 4.1

Check whether the following programmes are installed on your system, and if so, what version they are:

  1. curl
  2. wget
  3. efetch (after installing)
which curl && curl --version | head -n 1
which wget && wget --version | head -n 1
which efetch && efetch --version

curl is present on virtually all systems. wget is common on Linux and available in MobaXterm, but not installed by default on macOS. efetch should be present if the installation above succeeded.

Important🚀 Bonus: for those who want more

Look at what the Entrez Direct installer script actually does before running it:

curl -fsSL https://ftp.ncbi.nlm.nih.gov/entrez/entrezdirect/install-edirect.sh | less

Can you identify which directory it installs into, and which configuration file it modifies to update your PATH?

Tip🔑 Key points
  • which checks whether a programme is on your PATH; --version confirms what is installed
  • System package managers (apt, brew) are the standard way to install software on Linux and macOS; MobaXterm uses apt without sudo
  • Some tools are installed by downloading and running a shell script — only do this from trusted sources
  • NCBI Entrez Direct provides efetch and other tools for querying NCBI databases from the command line