Glossary

A beginner-friendly list of common terms you may see when working with code, GitHub, and servers.

A

API (Application Programming Interface)

A defined way for one tool to talk to another.
Example: a script can use a GitHub API to list files in a repository automatically.

Apptainer (formerly Singularity)

A container system often used on HPC (high‑performance computing) clusters.
Like Docker, it packages software and its dependencies, but it’s designed to work well on shared servers.

B

Branch

A separate “version line” of the code in Git.
You can make changes in a branch without affecting the main code until you merge it.

Build

The process of turning source code (and configuration) into something runnable, like an installed package or a container image.

C

CLI (Command Line Interface)

A way to interact with a computer by typing commands in a terminal instead of clicking buttons.

Cluster

A group of servers working together (common in research computing).

Commit

A saved checkpoint of changes in Git.
A commit includes what changed and usually a short message describing why.

Conda

A package and environment manager commonly used for Python (and other languages).
It helps you install dependencies and keep project environments separated.

Container

A packaged, isolated runtime environment that includes an application and everything it needs to run.
Containers help ensure “it runs the same everywhere”.

D

Dependency

A library/tool your code needs in order to run (for example: NumPy for Python).

Directory (Folder)

A container for files and other directories.
On most systems, “directory” and “folder” mean the same thing.

Docker

A popular container platform.
Docker images can package software so it runs the same on different machines, but some shared servers/HPC systems prefer Apptainer instead.

E

Environment (software environment)

A set of installed packages and settings used for a project.
Example: one environment for Project A and a different one for Project B, so they don’t conflict.

F

File path

The “address” of a file on a computer.
Example: data/raw/samples.csv (a file inside the data directory, then raw).

Fork

Your own copy of someone else’s GitHub repository under your account.
You can experiment safely and propose changes back via a pull request.

G

Git

A tool for tracking changes in files over time (version control).
Git helps you collaborate, revert mistakes, and understand project history.

GitHub

A website that hosts Git repositories and adds collaboration features (issues, pull requests, reviews).
Think of it as “Google Docs for code” (with extra tools).

Git repository (repo)

A project folder tracked by Git, including the full history of changes.

H

HPC (High‑Performance Computing)

Powerful shared computing resources used for large or long-running jobs (common in universities/research).

I

Issue

A GitHub item used to track work: bugs, questions, tasks, or feature requests.
Issues are great for documenting decisions and progress.

J

Job

A unit of work you run on a server or HPC system (often submitted to a scheduler).
Example: “run this analysis script for 4 hours using 8 CPUs”.

Job scheduler

Software that manages who gets to use which compute resources and when on a shared cluster.
Instead of everyone running heavy programs directly on login nodes, you submit jobs and the scheduler starts them when resources are available.

L

Linux

An operating system widely used on servers and HPC systems.
Many “server commands” are Linux commands.

Local machine

Your own laptop/desktop computer (as opposed to a remote server).

A server you connect to first on an HPC system (often via SSH).
You typically use it to prepare files and submit jobs, not to run heavy computations.

M

Main branch

The default primary branch of a repository (often called main or sometimes master).
It usually represents the most stable version of the project.

Merge

Combining changes from one branch into another (often into main).

N

Nextflow

A workflow manager for running pipelines (multi-step analyses) in a reproducible way.
It lets you define steps (processes) and how data flows between them, then runs them locally, on HPC (often with Slurm), or in the cloud.

Commonly used with:

Containers (Docker/Apptainer) to ensure the same software versions
nf-core: a community collection of well-tested Nextflow pipelines

O

Open source

Software whose source code is publicly available and can usually be used/modified under a license.

P

Partition (Slurm)

A named group of compute nodes in a Slurm cluster.
Partitions often represent different hardware or policies (e.g., cpu, gpu, long). You request a partition when submitting a job.

Pull request (PR)

A GitHub request to merge changes from one branch into another.
It’s a common way to propose changes, get review, and keep a record of discussion.

Q

Queue

A waiting line for jobs on a cluster.
Your job may wait in the queue until the requested resources are free.

R

README

A (usually) README.md file that explains what a repository is, how to install it, and how to use it.

Remote

A copy of your Git repository stored somewhere else (commonly on GitHub).
Your computer has a local copy; GitHub hosts a remote copy.

Repository hosting

A service (like GitHub) that stores Git repositories online.

S

sbatch (Slurm)

A Slurm command used to submit a batch job script to the scheduler.
You write a script describing resources needed (time, CPUs, memory) and what commands to run, then submit it with sbatch.

salloc (Slurm)

A Slurm command used to request an interactive allocation of resources.
Useful when you want a “live” session to run commands interactively on compute resources.

srun (Slurm)

A Slurm command used to run a program as part of a job/allocation.
Often used inside an sbatch script or after salloc.

scancel (Slurm)

A Slurm command used to cancel a queued or running job.

squeue (Slurm)

A Slurm command that shows jobs currently in the queue (waiting or running).

Slurm

A widely used job scheduler for HPC clusters.
It controls access to shared compute resources: you request resources (CPUs/GPUs/memory/time), submit a job, and Slurm runs it when resources are available.

Snakemake

A workflow manager (often used in bioinformatics/data science) to run analyses in multiple steps reliably.
You write rules like “to make file B, run this command using file A”, and Snakemake figures out what needs to run, in what order, and what can run in parallel.

Helpful features:

Automatically re-runs only steps whose inputs changed
Can run locally, on HPC (e.g., Slurm), or in the cloud
Works well with Conda and/or containers for reproducible software environments

Server

A computer that provides services to other computers (often accessed remotely).
In practice: “a computer you connect to over the network to run code or store data”.

Shell

The program that reads your terminal commands (e.g., bash, zsh).
It’s the “command interpreter”.

SSH (Secure Shell)

A secure way to connect to another computer (like a server) through the terminal.
Example: you can use SSH to log into a remote machine without physically being there.

T

Terminal

A text-based window where you type commands (especially common on Linux/macOS and servers).

Time limit / wall time (Slurm)

The maximum amount of real clock time your job is allowed to run.
If it exceeds this limit, it may be stopped automatically.

V

Version control

A system for tracking changes to files over time (Git is the most common one in software projects).

W

Workflow (GitHub Actions)

An automated set of steps that run on GitHub (tests, building docs, formatting checks, etc.).
Example: automatically run tests every time someone opens a pull request.

Workflow manager

Software that helps you run a multi-step analysis (a “pipeline”) in a reliable and reproducible way.

Instead of manually running step 1, then step 2, then step 3 (and trying to remember what you did), a workflow manager:

Defines each step and its inputs/outputs
Figures out the correct order automatically
Can run independent steps in parallel (faster)
Can restart and continue after interruptions
Often integrates with HPC schedulers (like Slurm) and containers (Docker/Apptainer)

Examples: Snakemake, Nextflow.