pipeline-cutandrun

ENCODE CUT&RUN Pipeline: FASTQ to Peaks and Signal Tracks

When to Use

• User wants to run a CUT&RUN or CUT&Tag processing pipeline from FASTQ to peaks
• User asks about "CUT&RUN pipeline", "CUT&Tag", "SEACR", "spike-in normalization", or "targeted chromatin"
• User needs to process CUT&RUN/CUT&Tag data with spike-in calibration and SEACR peak calling
• Example queries: "process my CUT&RUN FASTQs", "run SEACR on CUT&Tag data", "normalize CUT&RUN with spike-in controls"

Execute the CUT&RUN/CUT&Tag processing pipeline for targeted chromatin profiling, producing peak calls with SEACR and spike-in normalized signal tracks.

Pipeline Overview

FASTQ -> Trim -> Bowtie2 align (genome) -> Filter/dedup -> SEACR peaks
                     |                          |              |
              Bowtie2 align (spike-in)   Spike-in normalize  Signal tracks
                     |
              Scale factor calculation

ENCODE Repository

• GitHub: ENCODE-DCC/cutandrun-pipeline
• Container: encodedcc/cutandrun-pipeline
• This skill: Nextflow DSL2 reimplementation for portability

Core Tools and Versions

| Tool | Version | Purpose | Citation | |------|---------|---------|----------| | Bowtie2 | 2.5.3 | Alignment (genome + spike-in) | Langmead & Salzberg 2012 | | SEACR | 1.3 | Peak calling (CUT&RUN-specific) | Meers et al. 2019 | | MACS2 | 2.2.9.1 | Alternative peak caller | Zhang et al. 2008 | | Picard | 3.1.1 | Duplicate marking | Broad Institute | | samtools | 1.19 | BAM operations | Li et al. 2009 | | bedtools | 2.31.0 | Genomic arithmetic | Quinlan & Hall 2010 | | deepTools | 3.5.4 | Signal track generation | Ramirez et al. 2016 | | FastQC | 0.12.1 | Read quality | Andrews (Babraham) | | MultiQC | 1.21 | Aggregated QC | Ewels et al. 2016 |

Key Literature

1. Skene & Henikoff 2017 - "An efficient targeted nuclease strategy for

high-resolution mapping of DNA binding sites" (eLife, ~1,500 citations) DOI: 10.7554/eLife.21856

1. Meers et al. 2019 - "Peak calling by Sparse Enrichment Analysis for

CUT&RUN chromatin profiling" (Epigenetics & Chromatin, ~800 citations) DOI: 10.1186/s13072-019-0287-4

1. Kaya-Okur et al. 2019 - "CUT&Tag for efficient epigenomic profiling

of small samples and single cells" (Nature Communications, ~1,200 citations) DOI: 10.1038/s41467-019-09982-5

1. Nordin et al. 2023 - "The CUT&RUN suspect list of problematic regions"

(Genome Biology) DOI: 10.1186/s13059-023-02960-3

1. Amemiya et al. 2019 - "The ENCODE Blacklist" (Scientific Reports, ~1,372 citations)

DOI: 10.1038/s41598-019-45839-z

Execution

Quick Start (Local)

nextflow run main.nf \
    -profile local \
    --reads '/data/fastq/*_R{1,2}.fastq.gz' \
    --bowtie2_index '/ref/bowtie2_index/genome' \
    --spikein_index '/ref/bowtie2_ecoli/ecoli' \
    --chrom_sizes '/ref/hg38.chrom.sizes' \
    --blacklist '/ref/hg38-blacklist.v2.bed' \
    --outdir results/ \
    -resume

SLURM HPC

nextflow run main.nf \
    -profile slurm \
    --reads '/data/fastq/*_R{1,2}.fastq.gz' \
    --bowtie2_index '/ref/bowtie2_index/genome' \
    --spikein_index '/ref/bowtie2_ecoli/ecoli' \
    --chrom_sizes '/ref/hg38.chrom.sizes' \
    --blacklist '/ref/hg38-blacklist.v2.bed' \
    --outdir results/ \
    -resume

Cloud (GCP / AWS)

nextflow run main.nf \
    -profile gcp \
    --reads 'gs://bucket/fastq/*_R{1,2}.fastq.gz' \
    --bowtie2_index 'gs://bucket/ref/bowtie2_index/genome' \
    --spikein_index 'gs://bucket/ref/bowtie2_ecoli/ecoli' \
    --chrom_sizes 'gs://bucket/ref/hg38.chrom.sizes' \
    --blacklist 'gs://bucket/ref/hg38-blacklist.v2.bed' \
    --outdir 'gs://bucket/results/' \
    -resume

Resource Requirements

| Step | CPUs | RAM | Time (per sample) | |------|------|-----|-------------------| | Bowtie2 align (genome) | 8 | 8 GB | 30-60 min | | Bowtie2 align (spike-in) | 4 | 4 GB | 10-20 min | | Filter/dedup | 4 | 8 GB | 15-30 min | | SEACR peaks | 2 | 4 GB | 10-20 min | | Signal tracks | 4 | 8 GB | 15-30 min | | Total | 8 | 8 GB | 1.5-3 hours |

Pipeline Parameters

| Parameter | Default | Description | |-----------|---------|-------------| | --reads | required | Glob pattern to paired FASTQ files | | --bowtie2_index | required | Bowtie2 genome index prefix | | --spikein_index | required | Bowtie2 E. coli spike-in index prefix | | --chrom_sizes | required | Chromosome sizes file | | --blacklist | required | ENCODE blacklist BED file | | --outdir | ./results | Output directory | | --seacr_mode | stringent | SEACR mode: stringent or relaxed | | --seacr_norm | norm | SEACR normalization: norm or non | | --control | null | IgG control BAM (if available) | | --peak_caller | seacr | Peak caller: seacr or macs2 or both | | --skip_spikein | false | Skip spike-in normalization |

Output Files

results/
  fastqc/                           # Raw read quality
  alignment/
    {sample}.filtered.bam           # Filtered, deduplicated BAM
    {sample}.filtered.bam.bai
  spikein/
    {sample}.spikein_counts.txt     # Spike-in read counts
    {sample}.scale_factor.txt       # Computed scale factor
  peaks/
    {sample}.seacr.stringent.bed    # SEACR stringent peaks
    {sample}.seacr.relaxed.bed      # SEACR relaxed peaks
    {sample}.macs2_peaks.narrowPeak # MACS2 peaks (if requested)
  signal/
    {sample}.normalized.bw          # Spike-in normalized signal
    {sample}.fragments.bed          # Fragment BED file
  qc/
    {sample}.flagstat.txt
    {sample}.fragment_sizes.txt
    {sample}.frip.txt
  multiqc/
    multiqc_report.html

QC Thresholds

| Metric | Pass | Warning | Fail | |--------|------|---------|------| | Mapping rate (genome) | >80% | 60-80% | <60% | | Spike-in reads | 1-10% of total | 0.1-1% or 10-30% | <0.1% or >30% | | Duplication rate | <20% | 20-40% | >40% | | FRiP (peaks) | >10% | 5-10% | <5% | | Peak count | >5,000 | 1,000-5,000 | <1,000 | | Fragment size | Nucleosomal pattern | Irregular | No pattern |

Fragment Size Distribution

CUT&RUN produces a characteristic nucleosomal ladder:

• <120 bp: Sub-nucleosomal (TF binding)
• ~150 bp: Mononucleosomal (histone marks)
• ~300 bp: Dinucleosomal
• Absence of nucleosomal pattern suggests protocol issues

Spike-in Normalization

Spike-in normalization is CRITICAL for CUT&RUN quantitative comparison.

How It Works

1. E. coli DNA is carried over from pA-MNase/pA-Tn5 production
2. Each sample has a different amount of spike-in reads
3. Samples with more target cleavage have fewer spike-in reads (proportionally)
4. Scale factor = 1 / (spike-in reads / minimum spike-in reads across samples)

Scale Factor Calculation

Sample A: 200,000 spike-in reads -> scale = 1.0 (minimum)
Sample B: 400,000 spike-in reads -> scale = 0.5
Sample C: 100,000 spike-in reads -> scale = 2.0

Higher spike-in counts = less target enrichment = lower scale factor.

SEACR vs MACS2

| Feature | SEACR | MACS2 | |---------|-------|-------| | Designed for | CUT&RUN/CUT&Tag | ChIP-seq | | Background model | Sparse enrichment | Dynamic Poisson | | Control required | Optional (IgG) | Recommended | | Low background | Handles well | May overcall | | Stringent mode | Very conservative | Via q-value | | ENCODE recommendation | Primary for CUT&RUN | Alternative |

SEACR is specifically designed for the sparse, low-background signal profile of CUT&RUN data. MACS2 may overcall peaks due to the low background.

Critical Pitfalls

Spike-in Calibration is CRITICAL

Without spike-in normalization, quantitative comparisons between samples are unreliable. The amount of pA-MNase (or pA-Tn5) varies between experiments, and spike-in reads provide the internal calibration standard.

IgG Control vs No-Antibody Control

• IgG control: Non-specific antibody, captures background binding
• No-antibody: No antibody, c