ChIP-nf

A Nextflow pipeline for processing ChIP-seq data.

Installing Nextflow

Nextflow can be installed by using the following command:

curl -fsSL get.nextflow.io | bash

Running the pipeline

First you need to pull the pipeline using Nextflow:

$ nextflow pull guigolab/chip-nf
Checking guigolab/chip-nf ...
 downloaded from https://github.com/guigolab/chip-nf.git

You can get the pipeline help with the following command:

$ nextflow run chip-nf --help

N E X T F L O W  ~  version 0.24.1
Launching `guigolab/chip-nf` [nostalgic_franklin] - revision: 974a45c356 [master]

C H I P - N F ~ ChIP-seq Pipeline
---------------------------------
Run ChIP-seq analyses on a set of data.

Usage:
    chipseq-pipeline.nf --index TSV_FILE --genome GENOME_FILE [OPTION]...

Options:
    --help                              Show this message and exit.
    --index TSV_FILE                    Tab separted file containing information about the data.
    --genome GENOME_FILE                Reference genome file.
    --genome-index GENOME_INDEX_ FILE   Reference genome index file.
    --genome-size GENOME_SIZE           Reference genome size for MACS2 callpeaks. Must be one of
                                        MACS2 precomputed sizes: hs, mm, dm, ce. (Default: hs)
    --mismatches MISMATCHES             Sets the maximum number/percentage of mismatches allowed for a read (Default: 2).
    --multimaps MULTIMAPS               Sets the maximum number of mappings allowed for a read (Default: 10).
    --min-matched-bases BASES           Sets the minimum number/percentage of bases that have to match with the reference (Default: 0.80).
    --quality-threshold THRESHOLD       Sets the sequence quality threshold for a base to be considered as low-quality (Default: 26).
    --fragment-length LENGTH            Sets the fragment length globally for all samples (Default: 200).
    --remove-duplicates                 Remove duplicate alignments instead of just flagging them (Default: false).
    --rescale                           Rescale peak scores to conform to the format supported by the
                                        UCSC genome browser (score must be <1000) (Default: false).
    --shift                             Move fragments ends and apply global extsize in peak calling. (Default: false).

Input

The input data and metadata should be specified using a tab separated file and passing it to the pipeline command with the option --index. Here is an example of the file format:

sample1    sample1_run1     /path/to/sample1_run1.fastq.gz    -           H3
sample1    sample1_run2     /path/to/sample1_run2.fastq.gz    -           H3
sample1    sample1_run3     /path/to/sample1_run3.fastq.gz    -           H3
sample1    sample1_run4     /path/to/sample1_run4.fastq.gz    -           H3
sample2    sample2_run1     /path/to/sample2_run1.fastq.gz    control1    H3K4me2
control1   control1_run1    /path/to/control1_run1.fastq.gz   control1    input

The fields in the file correspond to:

1. identifier used for merging the BAM files

2. single run identifier

3. path to the fastq file to be processed

4. identifier of the input or - if no control is used

5. mark/histone or input if the line refers to a control

6. optional sample fragment length. If not specified the fragment length is estimated using SPP

Output

The pipeline will produce the following output data:

• Alignments

• pileupSignal, pileup signal tracks

• fcSignal, fold enrichment signal tracks

• pvalueSignal, -log_10(P) signal tracks

• narrowPeak, peak locations with peak summit, pvalue and qvalue (BED6+4)

• broadPeak, similar to narrowPeak (BED6+3)

• gappedPeak, both narrow and broad peaks (BED12+3)

Check MACS2 output files for details.

The output data information is written to a file called chipseq-pipeline.db created in the folder from where the pipeline is run. Here is an example of the db file:

sample1   /path/to/results/peakOut/sample1.pileup_signal.bw    H3         255     pileupSignal    0.9960   0.4393
sample1   /path/to/results/peakOut/sample1_peaks.narrowPeak    H3         255     narrowPeak      0.9960   0.4393
sample1   /path/to/results/sample1.bam                         H3         255     Alignments      0.9960   0.4393
sample1   /path/to/results/peakOut/sample1_peaks.gappedPeak    H3         255     gappedPeak      0.9960   0.4393
sample1   /path/to/results/peakOut/sample1_peaks.broadPeak     H3         255     broadPeak       0.9960   0.4393
sample2   /path/to/results/peakOut/sample2_peaks.gappedPeak    H3K4me2    200     gappedPeak      0.9995   0.7216
sample2   /path/to/results/peakOut/sample2.fc_signal.bw        H3K4me2    200     fcSignal        0.9995   0.7216
sample2   /path/to/results/peakOut/sample2.pval_signal.bw      H3K4me2    200     pvalueSignal    0.9995   0.7216
sample2   /path/to/results/peakOut/sample2_peaks.broadPeak     H3K4me2    200     broadPeak       0.9995   0.7216
sample2   /path/to/results/peakOut/sample2.pileup_signal.bw    H3K4me2    200     pileupSignal    0.9995   0.7216
sample2   /path/to/results/peakOut/sample2_peaks.narrowPeak    H3K4me2    200     narrowPeak      0.9995   0.7216
sample2   /path/to/results/sample2_GCCAAT_primary.bam          H3K4me2    200     Alignments      0.9995   0.7216

The fields in the file correspond to:

1. merge identifier

2. path

3. mark/histone

4. (estimated) fragment length

5. data type

6. NRF (Nonredundant Fraction)

7. FRiP (Fraction of Reads in Peaks)