Module 7 — Solutions

---

Exercise 1 — Your First Script

#!/bin/bash

echo "Alex"
date
echo "Ready to do bioinformatics."

Notes:

• date is a command, not a string — no quotes needed.
• The shebang must be the very first line, with no blank lines or spaces before it.
• Your name goes in the echo line — replace Alex with your own.

---

Exercise 2 — Variables and basename

#!/bin/bash

INPUT="Training/short_reads/paired/SRR1553607_2.fastq"
SAMPLE=$(basename $INPUT .fastq)

echo "Full path: $INPUT"
echo "Sample name: $SAMPLE"

Notes:

• $(basename $INPUT .fastq) runs basename with two arguments: the value of $INPUT and the extension .fastq to strip.
• The result is captured by $() and stored directly in SAMPLE.
• The echo lines mix literal text and variable values — Bash replaces $INPUT and $SAMPLE before printing.

---

Exercise 3 — A Simple Loop

#!/bin/bash

for FILE in Training/short_reads/paired/*.fastq
do
    SAMPLE=$(basename $FILE .fastq)
    echo "Found sample: $SAMPLE"
done

Notes:

• *.fastq matches all files ending in .fastq at that path. Bash expands it automatically.
• $FILE holds the full path on each pass; basename extracts the clean name.
• The output format Found sample: $SAMPLE matches what was asked for — check that your echo line uses exactly that text.

---

Exercise 4 — Log Files

#!/bin/bash

LOG="sample_list.txt"

echo "Sample list" > $LOG

for FILE in Training/short_reads/paired/*.fastq
do
    SAMPLE=$(basename $FILE .fastq)
    echo "$SAMPLE" >> $LOG
done

echo "Log saved to: sample_list.txt"

Notes:

• > $LOG creates a fresh file and writes the header line. If you accidentally used >> here, the header would be appended to whatever was already in the file from a previous run.
• >> $LOG inside the loop adds each sample name without overwriting the header or previous entries.
• The final echo goes to the terminal, not the log — no > or >> means the output prints to the screen as normal.

---

Exercise 5 — Capturing All Output with exec

#!/bin/bash

LOG_FILE="full_run.log"

exec > >(tee $LOG_FILE) 2>&1

echo "Run started"

for FILE in Training/short_reads/paired/*.fastq
do
    SAMPLE=$(basename $FILE .fastq)
    echo "Processing: $SAMPLE"
done

echo "Run finished"

Notes:

• exec > >(tee $LOG_FILE) 2>&1 must come before any commands whose output you want captured — placing it right after the variable declarations is the standard pattern.
• There are no >> or | tee on any individual line. The exec line handles everything.
• When you swap to exec 2> errors.log, the echo messages still print to the terminal normally — only error messages (from failed commands) would be written to errors.log. With no errors in this simple loop, errors.log will be empty or absent.

---

Exercise 6 — Build Your Own Pipeline Script

#!/bin/bash

DATA_DIR="Training/short_reads/paired"
OUTPUT_DIR="my_fastqc_output"
LOG_FILE="my_run_log.txt"

exec > >(tee $LOG_FILE) 2>&1

mkdir -p $OUTPUT_DIR

for FILE in $DATA_DIR/*.fastq
do
    SAMPLE=$(basename $FILE .fastq)
    echo "Quality checking: $SAMPLE"
    fastqc $FILE --outdir $OUTPUT_DIR
done

echo "Pipeline complete."

Notes:

• Using exec here means the script no longer needs any >> or | tee lines — all output, including any FastQC messages and errors, goes to both the terminal and $LOG_FILE automatically.
• echo "Pipeline complete." sits after done so it runs once, not once per file.
• Compare this to the Exercise 5 solution from the previous version that used manual >> — exec gives the same result with less effort and also captures errors that >> would have missed.