megSAP is a NGS data analysis pipeline for medical genetics, which is developed by the Institute of Medical Genetics and Applied Genomics and several collaborators from academia and industry. Since December 2016 the project is publicly available on GitHub, however closed-source development started already in 2012.
The design goals of the project are:
- state-of-the-art in terms of sensitivity/specificity,
- fast
- and usable for diagnostics:
- we use no tools that require a license for diagnostics
- extensive logging (tools, versions, parameters) ensures reproducability of results
- extensive testing before adding/updating tools or databases makes sure the results are valid
If you are interested to join the effort, please contact Marc Sturm.
The installaton of megSAP is quite time-consuming because many tools and big databases need to installed.
The installation instructions can be found here.
Documentation about the different analysis pipelines can be found here:
- DNA analysis (single sample)
- DNA analysis (multi-sample and trio)
- DNA analysis (somatic tumor-normal pair / tumor only)
- RNA analysis (expression)
Please report any issues or questions to the megSAP issue tracker.
- 16.07.2020: Added support for mapping on Illumina Dragen
- 16.07.2020: Added strand bias and allele balance to output VCF/GSvar files
- 25.06.2020: Added repeat expansion calling with ExpansionHunter (for exomes and genomes)
- 18.06.2020: Replaced freebayes by varscan2 in somatic tumor-only pipeline
- 19.05.2020: Updated databases (Clinvar 20200506, CADD 1.6, HGNC)
- 05.05.2020: Added Circos plots for exome/genome germline samples containing (CNVs, ROHs, BAFs)
- 24.04.2020: Added NGSD annotation of SV calls, which allows effective filtering.
- 03.03.2020: Implemented multi/trio CNV calling for shallow WGS
- 04.12.2019: Updated all tools and databases to the latest version
- 18.11.2019: Decoupled NGSD database and megSAP for annotations (data from NGSD is now exported to VCF/BED format and annotated just as any other data source)
- 07.11.2019: Enabled SV calling for exomes/panels
- 04.11.2019: Added option to automatically delete FASTQ files after mapping to save disk space (see settings.ini)
- 31.11.2019: Added support for shallow WGS as replacement for arrays
- 06.08.2019: Updated all tools and databases to the latest version
- 05.08.2019: Finished refactoring and update of CNV calling and annotation
- 02.05.2019: Genome data and annotation databases (about 210GB) are now always copied to the local /tmp/ folder. See
local_datainsettings.ini. - 05.04.2019: Using ClinCNV for exomes and genomes now (CnvHunter is still used for panels)
- 29.03.2019: Outdated gene names annotated by VEP are now corrected via the HGNC identifier if possible.
- 26.02.2019: Using gnomAD o/e instead of ExAC pLI in NGSD.
- 22.02.2019: Updated all tools and databases.
- 10.12.2018: Added germline CNV calling for WGS using ClinCNV
- 29.11.2018: WGS analysis now generates a single GSvar file with all relevant variants (Impact HIGH/MODERATE/LOW or AF<=2% in public databases)
- 22.11.2018: Retrieving CGI cancer type from new NGSD entry (somatic pipeline)
- 16.11.2018: Updated to freebayes 1.2.0 (parallelized to compensate for the increased runtime)
- 07.11.2018: Updated VEP from version 93.2 to 94.5
- 19.10.2018: Reverted to freebayes 1.1.0 because version 1.2.0 is too slow.
- 10.10.2018: Updated all tools and databases to the latest releases.
- 04.10.2018: Using Ensembl VEP for variant annotation now (was SnpEff)
- 07.08.2018: Removed most annotation from the 'filter' column and moved the functionality to GSvar.
- 20.07.2018: Updated ABRA2 version (attention: this changes indel positions - see ABRA2 2.06 changelog)
- 17.07.2018: Added '##PIPELINE' header line to GSvar files to keep track of the megSAP version the file was created with.
- 11.07.2018: Added UPD detection for trios.
- 22.04.2018: NGSD import of germline variants restricted to variants with AF<5% to improve performance.
- 20.04.2018: NGSD now handles analysis job queuing and execution on SGE.
- 13.03.2018: Refactoring of trio analysis: it is now based on multi-sample pipeline, produces a GSvar file with three sample columns, and calls off-target variants.
- 18.01.2018: Added b-allele frequency files for visualization in IGV.
- 08.01.2018: Updated tools (BWA, samtools, snpEff) and databases (ClinVar, gnomAD, HGMD).
- 15.12.2017: Added runs-of-homozygosity detection to the germline pipeline.