Tweet
Offering a user-paid NGS analysis pipeline in the cloud
Dear Community,
our lab has developed a NGS analysis pipeline for viral sequence analysis (both genomic resequencing and RNA-Seq, including some novel pathogen detection and viral quasi-species reconstruction approaches) that fulfills various analytical needs of our clinical cooperation partners. Since we believe that the pipeline might be of interest to the broader virological community we decided to integrate our workflow into Galaxy as a set of plugins.
However, we do need our in-house computational capacities for our own analyses and are targeting people who do not have time and experience to set up their own Galaxy installation (for example, clinical virologists). Therefore we would like to provide the finished product as a non-profit service in the cloud (for example, by using Amazon AWS, for which the Galaxy installation is pretty straightforward).
We envision (or at least, hope) that the pipeline will be used heavily and would therefore like to transfer the costs for the used resources to the user (e.g., the clinical virologist) in a pay-as-you-go manner. Amazon offers its 'DevPay' service for these purpose, basically an infrastructure to bill users of an AWS application instead of the application provider. Again, we do not want to make any profit with this service, we just want people to pay for the cpu cycles they burn with as few hassles for us as possible.
I could well imagine that other academic institutions have similar goals, e.g. being willing to offer their work for free without wanting to cover the costs for analysis resources. Does anyone have experience with similar kinds of setups and would like to offer advice? I am looking for hints about technical (e.g. how to best integrate a Cloud Galaxy with DevPay or similar) as well as legal hurdles that we might have to be careful about (e.g., are we in conflict with open-source software licenses if billing is involved, even if we do not make a profit etc.).
Thanks a bunch!
Sven
our lab has developed a NGS analysis pipeline for viral sequence analysis (both genomic resequencing and RNA-Seq, including some novel pathogen detection and viral quasi-species reconstruction approaches) that fulfills various analytical needs of our clinical cooperation partners. Since we believe that the pipeline might be of interest to the broader virological community we decided to integrate our workflow into Galaxy as a set of plugins.
However, we do need our in-house computational capacities for our own analyses and are targeting people who do not have time and experience to set up their own Galaxy installation (for example, clinical virologists). Therefore we would like to provide the finished product as a non-profit service in the cloud (for example, by using Amazon AWS, for which the Galaxy installation is pretty straightforward).
We envision (or at least, hope) that the pipeline will be used heavily and would therefore like to transfer the costs for the used resources to the user (e.g., the clinical virologist) in a pay-as-you-go manner. Amazon offers its 'DevPay' service for these purpose, basically an infrastructure to bill users of an AWS application instead of the application provider. Again, we do not want to make any profit with this service, we just want people to pay for the cpu cycles they burn with as few hassles for us as possible.
I could well imagine that other academic institutions have similar goals, e.g. being willing to offer their work for free without wanting to cover the costs for analysis resources. Does anyone have experience with similar kinds of setups and would like to offer advice? I am looking for hints about technical (e.g. how to best integrate a Cloud Galaxy with DevPay or similar) as well as legal hurdles that we might have to be careful about (e.g., are we in conflict with open-source software licenses if billing is involved, even if we do not make a profit etc.).
Thanks a bunch!
Sven