Bio Workflows Chipseq Pipeline

FASTQ files (IP + Input)
    |
    v
[1. QC & Trimming] -----> fastp
    |
    v
[2. Alignment] ---------> Bowtie2
    |
    v
[3. BAM Processing] ----> sort, markdup, filter
    |
    v
[4. Peak Calling] ------> MACS3
    |
    v
[5. QC] ----------------> FRiP, fingerprint plots
    |
    v
[6. Annotation] --------> ChIPseeker
    |
    v
Annotated peaks + QC report

# Process both IP and Input samples
for sample in IP_rep1 IP_rep2 Input_rep1 Input_rep2; do
    fastp -i ${sample}_R1.fastq.gz -I ${sample}_R2.fastq.gz \
        -o trimmed/${sample}_R1.fq.gz -O trimmed/${sample}_R2.fq.gz \
        --detect_adapter_for_pe \
        --qualified_quality_phred 20 \
        --length_required 25 \
        --html qc/${sample}_fastp.html
done

# Build index (once)
bowtie2-build genome.fa bt2_index/genome

# Align
for sample in IP_rep1 IP_rep2 Input_rep1 Input_rep2; do
    bowtie2 -p 8 -x bt2_index/genome \
        -1 trimmed/${sample}_R1.fq.gz \
        -2 trimmed/${sample}_R2.fq.gz \
        --no-mixed --no-discordant \
        --maxins 1000 \
        2> aligned/${sample}.log | \
    samtools view -@ 4 -bS -q 30 - | \
    samtools sort -@ 4 -o aligned/${sample}.bam
done

for sample in IP_rep1 IP_rep2 Input_rep1 Input_rep2; do
    # Mark and remove duplicates
    samtools fixmate -m aligned/${sample}.bam - | \
    samtools sort - | \
    samtools markdup -r - aligned/${sample}.dedup.bam

    # Index
    samtools index aligned/${sample}.dedup.bam

    # Remove chrM reads (high mitochondrial is common)
    samtools view -h aligned/${sample}.dedup.bam | \
        grep -v chrM | \
        samtools view -b - > aligned/${sample}.final.bam
    samtools index aligned/${sample}.final.bam
done

# Narrow peaks (TFs, sharp histone marks like H3K4me3)
macs3 callpeak \
    -t aligned/IP_rep1.final.bam aligned/IP_rep2.final.bam \
    -c aligned/Input_rep1.final.bam aligned/Input_rep2.final.bam \
    -f BAMPE \
    -g hs \
    -n experiment \
    --outdir peaks \
    -q 0.01

# Broad peaks (H3K27me3, H3K36me3)
macs3 callpeak \
    -t aligned/IP_rep1.final.bam aligned/IP_rep2.final.bam \
    -c aligned/Input_rep1.final.bam aligned/Input_rep2.final.bam \
    -f BAMPE \
    -g hs \
    -n experiment_broad \
    --outdir peaks \
    --broad \
    --broad-cutoff 0.1

# Calculate FRiP (Fraction of Reads in Peaks)
total_reads=$(samtools view -c aligned/IP_rep1.final.bam)
reads_in_peaks=$(bedtools intersect -a aligned/IP_rep1.final.bam -b peaks/experiment_peaks.narrowPeak -u | samtools view -c)
frip=$(echo "scale=4; $reads_in_peaks / $total_reads" | bc)
echo "FRiP: $frip"

# Generate bigWig for visualization
bamCoverage -b aligned/IP_rep1.final.bam \
    -o bigwig/IP_rep1.bw \
    --normalizeUsing RPKM \
    -p 8

# Fingerprint plot (assess enrichment)
plotFingerprint \
    -b aligned/IP_rep1.final.bam aligned/Input_rep1.final.bam \
    --labels IP Input \
    -o qc/fingerprint.pdf

library(ChIPseeker)
library(GenomicFeatures)
library(rtracklayer)

# Custom GTF approach (preferred when a GTF is provided)
txdb <- makeTxDbFromGFF('annotation.gtf', format = 'gtf')
# Standard genome approach (when no custom GTF)
# library(TxDb.Hsapiens.UCSC.hg38.knownGene)
# txdb <- TxDb.Hsapiens.UCSC.hg38.knownGene

peaks <- readPeakFile('peaks/experiment_peaks.narrowPeak')
# overlap='all' couples gene assignment with feature overlap (host-gene convention);
# default overlap='TSS' assigns nearest-TSS gene independently of feature overlap
peak_anno <- annotatePeak(peaks, TxDb = txdb, tssRegion = c(-2000, 2000), overlap = 'all')

# Map gene symbols from GTF (annoDb only works with pre-built TxDb)
gtf <- import('annotation.gtf')
gene_map <- unique(data.frame(
    gene_id = sub('\\..*', '', gtf$gene_id),
    symbol = gtf$gene_name, stringsAsFactors = FALSE))
anno_df <- as.data.frame(peak_anno)
anno_df$geneId_base <- sub('\\..*', '', anno_df$geneId)
anno_df$SYMBOL <- gene_map$symbol[match(anno_df$geneId_base, gene_map$gene_id)]

plotAnnoPie(peak_anno)
plotDistToTSS(peak_anno)

write.csv(anno_df, 'peaks/annotated_peaks.csv', row.names = FALSE)
promoter_genes <- unique(anno_df$SYMBOL[grepl('Promoter', anno_df$annotation)])
write.table(promoter_genes, 'peaks/promoter_genes.txt', row.names = FALSE, col.names = FALSE, quote = FALSE)

#!/bin/bash
set -e

THREADS=8
GENOME="genome.fa"
INDEX="bt2_index/genome"
IP_SAMPLES="IP_rep1 IP_rep2"
INPUT_SAMPLES="Input_rep1 Input_rep2"
OUTDIR="results"

mkdir -p ${OUTDIR}/{trimmed,aligned,peaks,qc,bigwig}

# Step 1: QC
for sample in $IP_SAMPLES $INPUT_SAMPLES; do
    fastp -i ${sample}_R1.fastq.gz -I ${sample}_R2.fastq.gz \
        -o ${OUTDIR}/trimmed/${sample}_R1.fq.gz \
        -O ${OUTDIR}/trimmed/${sample}_R2.fq.gz \
        --html ${OUTDIR}/qc/${sample}_fastp.html -w ${THREADS}
done

# Step 2-3: Align and process
for sample in $IP_SAMPLES $INPUT_SAMPLES; do
    bowtie2 -p ${THREADS} -x ${INDEX} \
        -1 ${OUTDIR}/trimmed/${sample}_R1.fq.gz \
        -2 ${OUTDIR}/trimmed/${sample}_R2.fq.gz \
        --no-mixed --no-discordant 2> ${OUTDIR}/qc/${sample}_align.log | \
    samtools view -@ ${THREADS} -bS -q 30 - | \
    samtools fixmate -m - - | \
    samtools sort -@ ${THREADS} - | \
    samtools markdup -r - ${OUTDIR}/aligned/${sample}.bam
    samtools index ${OUTDIR}/aligned/${sample}.bam
done

# Step 4: Peak calling
ip_bams=$(for s in $IP_SAMPLES; do echo "${OUTDIR}/aligned/${s}.bam"; done | tr '\n' ' ')
input_bams=$(for s in $INPUT_SAMPLES; do echo "${OUTDIR}/aligned/${s}.bam"; done | tr '\n' ' ')

macs3 callpeak -t ${ip_bams} -c ${input_bams} \
    -f BAMPE -g hs -n experiment \
    --outdir ${OUTDIR}/peaks -q 0.01

echo "Pipeline complete. Peaks: ${OUTDIR}/peaks/experiment_peaks.narrowPeak"