COSMOS: Workflow Management System
COSMOS is a Python library for workflow management that allows formal description of pipelines and partitioning of jobs. In addition, it includes a user-interface for tracking the progress of jobs, abstraction of the queuing system and fine-grained control over the workflow. Workflows can be created on traditional computing clusters as well as cloud-based services. It is developed jointly by the Laboratory for Personalized Medicine at Harvard Medical School, Invitae, a clinical next generation sequencing diagnostic laboratory, and the Wall Lab at Stanford University.
How to Cite COSMOS
Gafni E, Luquette LJ, Lancaster AK, Hawkins JB, Jung J-Y, Souilmi Y, Wall DP, Tonellato PJ: COSMOS: Python library for massively parallel workflows. Bioinformatics (2014) 30 (20): 2956-2958. doi: 10.1093/bioinformatics/btu385 |
Download COSMOS Code
COSMOS is avaialable in two versions:
COSMOS 2.0
COSMOS 2.0 is production ready and available for download fully open source under an MIT license
COSMOS 2.0 code is available on gihtub as well as through Python PIP repository.COSMOS 1.0 (deprecated)
COSMOS 1.0 is available for download FOR ACADEMIC NON-COMMERCIAL RESEARCH PURPOSES.
COSMOS Documentation
COSMOS documentations for the version 1.0 and 2.0 are available:
COSMOS 2.0
COSMOS 2.0 release documentation is available to build from the code, and also accessible online here: Access online documentation
COSMOS 1.0 (deprecated)
COSMOS 1.0 documentation is available to build from the code, and also accessible here: Access online documentation
COSMOS Publications
BMC Medical Genomics
Souilmi Y., Lancaster AK, Jung J-Y., Ettore R., Jared HB., Ryan Powles, Amzazi S., Ghazal H., Tonellato, P., Wall DP. (2015). Scalable and Cost-effective NGS Genotyping in the Cloud. Forthcoming.
BMC Bioinformatics: ISCB Student Council Symposium 2014 Highlight
Souilmi Y., Jung J-Y., Lancaster AK, Gafni E., Amzazi S., Ghazal H., Wall DP., Tonellato, P. (2015). COSMOS: cloud enabled NGS analysis. BMC Bioinformatics, 16(Suppl 2), A2. doi: 10.1186/1471-2105-16-S2-A2
Bioinformatics Application Note
Gafni E, Luquette LJ, Lancaster AK, Hawkins JB, Jung J-Y, Souilmi Y, Wall DP, Tonellato PJ: COSMOS: Python library for massively parallel workflows. Bioinformatics (2014) 30 (20): 2956-2958. doi: 10.1093/bioinformatics/btu385
COSMOS Compatible Workflows
We share here COSMOS compatible workflows:GenomeKey
GenomeKey 1.0 is a COSMOS 1.0 implementation of GATK best practices. It offers high-performance and high-throughput whole exome/genome data analysis. GenomeKey's performance was systematically benchmarked as shown in the BMC Bioinformatics preview and the full BMC Medical Genomics paper.
GenomeKey 1.0 is available under APACHE 2.0 LICENCE.
Note 1: We are also building a COSMOS 2.0 compatible version.
Note 2: For instructions on how to use Genomekey on AWS, please refer to the github wiki page.
PvKey
PvKey 1.0 is a COSMOS 1.0 workflow for Tumor-Normal matched samples analysis. It calls somatic variants using Mutect and structural variants using SVDetect. It can handle genome, exome and targeted samples (TruSeq Custom Amplicon).
PvKey 1.0 is available under APACHE 2.0 LICENCE.
Note 1: We are also building a COSMOS 2.0 compatible version.
Note 2: For instructions on how to use PvKey, please refer to the github instructions page.
Where can I find documentation?
You can refer to the documentation section for more details.
FAQ
Is COSMOS free to use?
COSMOS is available in two versions 1.0 and 2.0. Version 1.0 is free and open FOR ACADEMIC NON-COMMERCIAL RESEARCH PURPOSES only, please check the license agreement. COSMOS 2.0 is production ready, open source and available under MIT LICENCE.
How can I get COSMOS?
You can refer to the download section for more details.
Where can I find documentation?
You can refer to the documentation section for more details.
Where can I find support?
You use the Please use the Github Issue Tracker, or contact us on the support room: 
Is COSMOS 2.0 compatible with 1.0's workflows?
COSMOS 2.0 was built with the same philosophy and keeps the same feature and adds many more. The new API was built using modern technics and is not backwards-compatible with 1.0 pipelines. However, the porting is not very complicated once you familiarize yourself with the new syntax, which is more self explenatory. Please refer to Access online documentation for more details.