1. Epigenomic Alignment vs RNA-Seq Alignment

Unlike transcriptomic RNA-seq, which requires splice-aware aligners (such as STAR or HISAT2) to bridge exon-exon junctions across large introns, ChIP-seq fragments originate directly from continuous genomic DNA. Therefore, splice-unaware, fast, index-based aligners such as Bowtie2 or BWA-MEM2 are mandatory.

2. Building the GRCh38 Genome Index & Running Bowtie2

# Download GRCh38 major chromosomes reference and build index
bowtie2-build --threads 8 GRCh38.primary_assembly.genome.fa grch38_idx

# Align trimmed reads with end-to-end sensitive settings
for sample in HD_Rep1 HD_Rep2 Control_Rep1 Control_Rep2 HD_Input Control_Input; do
    echo "Aligning ${sample} with Bowtie2..."
    bowtie2 \
        -p 8 \
        -x grch38_idx \
        -U 01_trimmed/${sample}.trimmed.fastq.gz \
        --very-sensitive \
        --no-unal \
        2> 06_qc_reports/${sample}_bowtie2.log \
        | samtools view -bS - \
        | samtools sort -@ 4 -m 2G -o 02_aligned/${sample}.sorted.bam -
    
    samtools index 02_aligned/${sample}.sorted.bam
done

3. Post-Alignment Quality Filtering (MAPQ >= 30)

Repetitive genomic elements (LINEs, SINEs, segmental duplications) generate ambiguous multi-mapping reads that can create artificial peak artifacts. We discard reads with Mapping Quality scores lower than 30 and remove unmapped or supplementary alignments using SAM flags:

for sample in HD_Rep1 HD_Rep2 Control_Rep1 Control_Rep2 HD_Input Control_Input; do
    # -q 30: Minimum mapping quality score of 30
    # -F 1804: Exclude unmapped, secondary, vendor-failed, and PCR duplicate reads
    samtools view -@ 4 -b -q 30 -F 1804 \
        02_aligned/${sample}.sorted.bam > 03_filtered/${sample}.q30.bam
    
    samtools index 03_filtered/${sample}.q30.bam
    
    echo "Alignment flag statistics for ${sample}:"
    samtools flagstat 03_filtered/${sample}.q30.bam
done