This pipeline calls peaks in ChIP-Seq data with MACS2 and DFilter and performs motif discovery with MEME-ChIP.
The following steps are performed:
- Read mapping with BWA aln by default. See
cfg/references.yaml
for default references and also refer to option--references-cfg
- Filtering of duplicated, unmapped reads and reads with bad mapping quality (Q20).
- By default the pipeline will call peaks with both, MACS and DFilter. This behaviour can be changed with "--skip-macs2" and "--skip-dfilter".
- Extraction of sequences around extended peak regions followed by motif discovery with MEME-ChIP
Note, running a ChIPseq analysis requires a control and treatment
sample. When using a sample config as input (--sample-cfg
or -S
),
make sure that the control is named control
.
The peak type is library dependent and can be set with --peak-type
or -t
. If set to TF
or histone-narrow
, MACS2 will call "narrow
peaks" in either case and DFilter options will be adjusted accordingly
If set to histone-broad
, MACS2 will call "broad peaks" and "narrow
peaks" and DFilter options will be adjusted accordingly. See the
DFilter
tutorial
for an explanation of options.
Processed BAM files (*.bwa-aln-nsrt.bam
) for control and treatment
samples can be found in the correspondingly named subfolders in
./out/
.
Peak calling results, motif discovery results and mapping stats can be
found in the ./out/treatment
subfolders. The most important are
treatment_peaks.xls
: Called peaks and peaks coordinates (MACS2 only)treatment.Peaks
: Called peaks and peaks coordinates (DFilter only)treatment_summits.bed
: Peak summit locations for every peaktreatment_control_lambda.bw
: Control tag density profiletreatment_treat_pileup.bw
: Treatment tag density profiletreatment_filtered_treatment.wig
: Background smoothened treatment tag density profile (DFilter only)treatment-memechip/index.html
: Motif discovery resultstreatment.bwa-aln-nsrt.bamstats/stats.txt
: Mapping stats