megSAP - DNA long-read analysis (single sample)

Basics

Single sample long-read DNA analysis is performed using the analyze_longread.php script.
Please have a look at the help using:

> php megSAP/src/Pipelines/analyze_longread.php --help

The main parameters that you have to provide are:

• folder - The sample folder, which contains the raw reads including methylation information in unmapped BAM format.
• name - The sample name, which must be the basename of the BAM file.
• steps - Analysis steps to perform. For example ma,vc,sv,ph,an to perform mapping, small variant calling and structural variant calling (with annotation).
• system - The processing system INI file.

Running an analysis

The analysis pipeline assumes that that the raw read data in unmapped BAM format is in the sample folder. If that is the case, the whole analysis is performed with one command, for example like this:

php megSAP/src/Pipelines/analyze_longread.php -folder Sample_NA12878_01 -name NA12878_01 -system SQK-114.ini -steps ma,vc,cn,sv,re,an

In the example above, the configuration of the pipeline is done using the SQK-114.ini file, which contains all necessary information (see processing system INI file).

Tools used in this analysis pipeline

The following tools are used for mapping, methylation analysis and small variant calling/annotation (steps ma,vc,me,an):

step	tool	comments
mapping	minimap2
mapping - extract methylation info	modkit
variant calling - calling of SNVs and InDels	DeepVariant
variant calling - decompose complex variants	vcfallelicprimitives
variant calling - break multi-allelic variants	VcfBreakMulti
variant calling - left-normalization of InDels	VcfLeftNormalize
annotation - general	VcfAnnotateFrom*

CNV calling/annotation (steps cn,an) is performed using these tools:

step	tool	comments
CNV calling	ClinCNV
annotation - general	BedAnnotateFromBed	Several data sources are annotated using this tool.
annotation - gene information	CnvGeneAnnotation
annotation - overlapping pathogenic CNVs from NGSD	NGSDAnnotateCNV

SV calling/annotation (steps sv,an) is performed using these tools:

step	tool	comments
SV calling	Sniffles
annotation - gene information	BedpeGeneAnnotation
annotation - matching SVs from NGSD	BedpeAnnotateCounts
annotation - breakpoint density from NGSD	BedpeAnnotateBreakpointDensity

RE calling (step re) is performed using these tools:

step	tool	comments
RE calling	Straglr

Phasing (step ph) is performed using these tools:

step	tool	comments
phasing	longphase

A complete list of all tools and databases used in megSAP including version and when they were last updated can be found here.

Performance

Performance benchmarks of the the megSAP pipeline can be found here.

Output

After the analysis with the steps ma,vc,sv,an,me, these files are created in the output folder:

1. mapped reads in CRAM format
2. small variant list in VCF format
3. small variant list in GSvar format
4. structural variant list in VCF format
5. methylation information in TSV/VCF format
6. QC data in qcML format, which can be opened with with GSvar or Firefox