local to cloud data transfer, and user data stored on cloud Konstantinos Krampis, JCVI 28 November 2011 It would be of benefit to the readers if the authors could provide a data point for the time required for data transfer of user input data from the local CLOVR virtual machine to the cloud. In addition, they could consider in a future version to permanently store the user-uploaded input dataset on the cloud (S3 for example) along with the reference databases. The reason is that users might require to re-run multiple times a pipeline with the same input dataset, but using different parameters (consider days in between runs, so the virtual cluster has to be turned on and off each time). The user input data can be retrieved each time from the cloud storage using the cloud credentials provided by the user through the CLOVR interface. Uploading the same input dataset before each run to the cloud through the CLOVR virtual machine creates additional time-delay overhead, and also results in multiple data-in transfer fees charged by the cloud provider. Competing interests No competing interests declared.