README for ASAP v2.1

Introduction

ASAP is a flexible bioinformatic pipeline for ATAC-seq data analysis. Starting from raw ATAC-seq sequencing reads, ASAP outputs raw and filtered mapping files, coverage files (reads coverage ; tn5 insertion events coverage), fragment length distribution, read exraction based on fragment length, and peak calling results.

Overview of major steps

• Mapping
• Post-mapping processing and filtering:
  • Filter (or not) reads that fall into user-defined blacklisted regions
  • Select reads that do not carry more than minMismatch and filter by minimum mapping quality (MAPQ)
  • Mark duplicated pairs
  • Select concordant, non-duplicated pairs.
  • Shift reads by 4bp as described in Schep et al.,2015: shift by 4bp toward to center of the transposition event.
• Compute read coverage
• Compute insertion events coverage
• Fragment length distribution
• Extract reads pairs based on a fragment length range and compute arcs between fragment extremities (protection visualization)
• Peak calling

ASAP is:

User-friendly: requires a single configuration file. Thus, only one option is required when running the command line (see Usage of ASAP)

Flexible: provides the possibility to skip a given step(s) and target specific post-processing step(s).

Dependencies

• Bowtie2 version >= 2.2.9
• MACS2
• Samtools version >=1.2
• GenomeCoverageBed version >= v2.20.1

Usage of ASAP

A configuration file required to execute the pipeline.

bash ASAP.sh [-h] [-v] [-c]

Options

-c CONFIGFILE

This is the only REQUIRED parameter for ASAP. The configuration file is a text file that gathers the full set of parameters required to execute the pipeline. (check the example ASAP_configFile_example.conf in distribution)

-h/-v

Print out the help/current version

About the configuration file:

The configuration file gathers the parameters of each step. Note that, when running the pipeline, only the "turned on" steps will performed. A step is turned on by a yes/no argument.

Here we list the different set of parameters to be filled in the configuration file:

General parameters

General information option about the run. Must be always filled.

OUTDIR:	Main output directory where results are written. OUTDIR is created if does not exist
sampleName:	Name of the processed sample. No space is allowed: use _ or - to mimic space if needed
CHRLEN:	Chromosome info file (tab-delimited format: <Chr name><chr length>)
path:	Full path to the different dependencies, if not already added to $PATH

Mapping step parameters

map:	Set to "yes/no". If mapping is skipped (map=no), a BAM file must be provided to proceed. (see post-mapping steps).
FASTQ1:	fastq file (R1). File can be gizpped
FASTQ2:	fastq file (R2). File can be gizpped
bowtieIndex:	Prefix of bowtie2 indexes
mappingParameters:	Bowtie2 mapping parameters. Default: --very-sensitive -X 2000 -p 10

Post mapping steps

It is possible to skip the mapping step (map=no) and perform any of the post-mapping steps. To do so, aligned reads must be provided in a BAM file. If* map=yes*, the "turned on" post-mapping steps will performed on the internal mapping results.

BAM:	aligment file in BAM format

Filtering parameters

filter:	Set to "yes/no". If map=yes, filtering will be performed on internal mapping results, if map=no, filtering will be performed on the provided alignment file in BAM option.
maxMis:	Maximum number of mismatches allowed per read.
blacklist:	Set to "yes/no" if reads should be filtred based on a list of blacklisted regions. If "blacklist=yes", blacklisted regions must be provided in the next parameter.
blacklistedRegions:	Regions used to filter reads.(tab-delimited format: <Chr name><start><end>)
shift:	Set to "yes"/"no". If shift=yes, reads are shifted by 4bp so that read starts reflect the center of the Tn5 transposition event

Coverage

readCoverage:	Set to "yes/no" if read coverage should be computed or not
ieventsCoverage:	Set to "yes/no" if Tn5 insertion events coverage should be computed or not

Read extraction

extractReads:	Set to "yes/no" if read pairs should be extracted based on a given range of fragment length
lowBoundary:	Lower boundery of the range: [lowBoundary,upBoundary]. Default=100
upBoundary:	Upper boundery of the range: [lowBoundary,upBoundary]. Default=250
arcs:	set to "yes/no" if extracted fragments should represented as arcs (linked extremities)

Fragment length

fragDist:	Set to "yes/no" if fragment length distribution should be computed or not

Peak calling

callpeak:	Set to "yes/no" if peak calling should be computed or not.
control:	Control bam file. Note that peak calling can be performed without a control, however, one can provide a control such as ATAC-seq on genomic DNA. Leave option empty if no control is used.
MODE:	Peak calling mode: <broad/narrow>. Default=broad
modelParameters:	MACS2 shifting options
fdr:	Cutoff for peak detection. Default=0.01
gsize:	Effective genome size of tair10 (gsize=10e7)

Output files

ASAP outputs mapping files, coverage files, fragments distribution table/plot and MACS2 peak calling results.

Mapping output

*.mapped.sorted.bam:	Contains mapped reads (bowtie2 raw mapping results)

Filtering/post-processing outputs

*.(un)masked.(un)shifted.bam:	Contains the selected set of reads after filtering. Ideally, accessible peaks are called using this file.
*.csv:	Summary of filtering step is CSV format

Coverage outputs

*.(un)masked.(un)shifted.ievent.bam:	Contains Tn5 insertion events. Basically, instead of showing reads, only the position corresponding to Tn5 insertion event are shown)
*.(un)masked.(un)shifted.bw:	Genome-wide coverage of ATAC reads
*.(un)masked.(un)shifted.ievent.bw:	Genome-wide coverage of Tn5 insertion events

Read extraction

*.subReads.f3.frag*.bam:	Contains the set of extracted reads based on the given rage of fragment length
*.subReads.f3.frag*.bw:	Genome-wide coverage of the set of extracted reads based on the given rage of fragment length
*.subReads.f3.frag*.arcs.bed:	arcs between fragment extremities. This file is visualized on IGV

Fragment length distribution

*.TLEN.f3F16.txt:	Counts/frequencies of fragments length
*.TLEN.f3F16.png:	Plot of fragment length distribution

Peak calling outputs

Output are stored in an directory: peak_calling_<sampleName>. Check MACS2 output list