2: Filesystem Hierarchy & Directory Operations

Chapter 2: Filesystem Hierarchy & Directory Operations

Unlike Windows, which assigns distinct drive letters (e.g., C:\, D:\) to physical disks, Linux organizes all storage devices into a single unified tree starting at the root directory (/). Understanding this directory hierarchy and mastering rapid path traversal is essential for organizing complex, multi-sample sequencing cohorts.

1. The Filesystem Hierarchy Standard (FHS)

• / : The Root directory. The parent of all directories on the system.
• /home/username : The user's personal home directory. Contains shell configuration files (.bashrc) and private documents. Abbreviated as ~.
• /etc : System-wide configuration files (networking, user accounts).
• /tmp : Ephemeral scratch space. Cleaned automatically upon reboot.
• /opt : Third-party software packages and standalone bioinformatics binaries.
• /var/log : Operating system and service log files.

2. Absolute vs. Relative Paths

Paths define the location of a file within the filesystem hierarchy:

• Absolute Path: Starts from the root directory (/). Always points to the exact same file regardless of your current location (e.g., /scratch/genomics/project1/sample_01.fastq.gz).
• Relative Path: Defined relative to your current working directory (e.g., ./data/sample_01.fastq.gz or ../reference/genome.fa).

# Special path symbols:
.   # Current working directory
..  # Parent directory (one level up)
~   # Current user's home directory
-   # Previous working directory (toggle back)

3. Directory Navigation & Inspection

# Print the current working directory
pwd

# Change into project directory
cd /scratch/genomics/project1

# Jump up one directory level
cd ..

# Jump back to your personal home directory
cd ~
# or simply:
cd

# Toggle back to the previous directory you were in
cd -

# List files with detailed metadata (permissions, owner, size, timestamp)
ls -la

# List files sorted by modification time (most recent at the bottom)
ls -lhtr

# Visualize directory tree structure up to 2 levels deep
tree -L 2

4. Directory Creation, Copying & Deletion

# Create nested project directories in a single command with -p (parents)
mkdir -p rnaseq_cohort/{raw_fastq,trimmed,alignments,feature_counts,reports}

# Copy files (use -r or -R for recursive directory copying)
cp sample1_R1.fastq.gz backup/
cp -r rnaseq_cohort/ rnaseq_cohort_backup/

# Move or rename files
mv old_name.fastq.gz sample01_R1.fastq.gz
mv *.bam alignments/

# Deletion commands (USE WITH EXTREME CAUTION - Linux rm has no Recycle Bin!)
rm temp_file.txt
rm -i cautious_delete.bam  # Prompt before each deletion
rm -rf temporary_run/      # Force recursive deletion without confirmation

5. Wildcard Expansion (Globbing) for NGS Samples

Bash shell wildcards allow you to batch-process dozens of sample files simultaneously without writing loops:

# Asterisk (*): Matches zero or more characters
ls *.fastq.gz          # All compressed FASTQ files
ls sample_*_R1.fq.gz   # Forward reads for all samples

# Question mark (?): Matches exactly one character
ls sample_??.bam       # Matches sample_01.bam, sample_12.bam, but not sample_100.bam

# Character classes ([]): Matches any character enclosed
ls sample_[1-5]_R1.fastq.gz

# Brace expansion: Generates arbitrary combinations
touch sample_{A,B,C}_{R1,R2}.fastq.gz
# Creates: sample_A_R1, sample_A_R2, sample_B_R1, sample_B_R2, sample_C_R1, sample_C_R2