Preprocessing & Quality Control

Step 1: Assessing read quality and adapter trimming

Overview

Quality control is the critical first step in any RNA-seq analysis. Poor quality data can lead to unreliable downstream results. This step involves:

1. Quality assessment of raw reads
2. Adapter trimming and quality filtering
3. Post-trimming QC to verify improvement

Quality Assessment with FastQC

FastQC provides a comprehensive overview of various quality metrics for each sample.

#| eval: false
#| code-fold: false

# Run FastQC on all FASTQ files
fastqc -t 8 -o fastqc_results/ raw_data/*.fastq.gz

# Aggregate results with MultiQC
multiqc fastqc_results/ -o multiqc_report/

Key Metrics to Evaluate

NoteQuality Indicators

• Per base sequence quality: Should be >28 across most positions
• Per sequence quality scores: Peak should be >30
• Adapter content: Should be minimal (<5%)
• Overrepresented sequences: Check for contaminants

Adapter Trimming with fastp

fastp is a fast all-in-one preprocessing tool that performs quality filtering, adapter trimming, and quality control.

#| eval: false
#| code-fold: false

#!/bin/bash
# Adapter trimming with fastp

THREADS=8
INPUT_DIR="raw_data"
OUTPUT_DIR="trimmed_data"

mkdir -p ${OUTPUT_DIR}

for R1 in ${INPUT_DIR}/*_R1.fastq.gz; do
    SAMPLE=$(basename ${R1} _R1.fastq.gz)
    R2=${INPUT_DIR}/${SAMPLE}_R2.fastq.gz
    
    fastp \
        --in1 ${R1} \
        --in2 ${R2} \
        --out1 ${OUTPUT_DIR}/${SAMPLE}_R1_trimmed.fastq.gz \
        --out2 ${OUTPUT_DIR}/${SAMPLE}_R2_trimmed.fastq.gz \
        --html ${OUTPUT_DIR}/${SAMPLE}_fastp.html \
        --json ${OUTPUT_DIR}/${SAMPLE}_fastp.json \
        --thread ${THREADS} \
        --qualified_quality_phred 20 \
        --length_required 36 \
        --detect_adapter_for_pe \
        --correction
        
    echo "Finished processing: ${SAMPLE}"
done

fastp Parameters Explained

Parameter	Description	Recommended Value
--qualified_quality_phred	Minimum base quality	20
--length_required	Minimum read length after trimming	36
--detect_adapter_for_pe	Auto-detect adapters for PE reads	Enable
--correction	Enable base correction for PE overlap	Enable

Alternative: Trimmomatic

#| eval: false
#| code-fold: false

# Trimmomatic example
java -jar trimmomatic-0.39.jar PE \
    -threads 8 \
    input_R1.fastq.gz input_R2.fastq.gz \
    output_R1_paired.fastq.gz output_R1_unpaired.fastq.gz \
    output_R2_paired.fastq.gz output_R2_unpaired.fastq.gz \
    ILLUMINACLIP:adapters.fa:2:30:10:2:keepBothReads \
    LEADING:3 \
    TRAILING:3 \
    SLIDINGWINDOW:4:15 \
    MINLEN:36

Post-Trimming QC

After trimming, run FastQC again to verify quality improvement:

#| eval: false

# Post-trimming QC
fastqc -t 8 -o fastqc_trimmed/ trimmed_data/*.fastq.gz
multiqc fastqc_trimmed/ fastp_reports/ -o multiqc_final/

TipBest Practice

Always compare pre- and post-trimming metrics. You should see:

• Improved per-base quality scores
• Reduced adapter content
• Slight reduction in total reads (filtered low-quality reads)

Troubleshooting

WarningCommon Issues

1. High adapter content: Check if correct adapter sequences are being used
2. Low quality at 3’ end: Normal for Illumina; trimming should resolve this
3. Overrepresented sequences: May indicate contamination or rRNA

Next Steps

Once QC is complete and reads are trimmed, proceed to Read Alignment.