Abstract : Bioinformatics applications are often structured as workflows that are composed of a set of operations to perform on large data sets. These workflows are deployed as complex scripts that handle the sequence of program calls with their relevant inputs and try to take advantage of a computer cluster using a scheduler. Their performances rely on the user ability to analyze the potential parallelism in the workflow. SLICEE (Service Layer for Intensive Computation Execution Environment) abstracts the scheduler cluster calls by handling command submission, parallelism extraction and data management. A workflow client orchestrates the SLICEE services that exploit the data parallelism, and takes care of the data routing between tasks. Thus the workflow tasks execution takes advantage of the parallelism available on the cluster with minimum user intervention. Maintaining a cluster architecture is expensive and its processing power is hard to scale over time. Cloud computing proposes to virtualize a computer architecture and to deploy it on available physical computing resources. Therefore the physical architecture is shared and the virtual processing power can be scaled to meet the user demand. OBIWEE (On Demand Bioinformatics Intensive Workflow Execution Environment) is a bioinformatics intensive workflow execution environment preconfigured on a linux virtual cluster, that can be deployed either on a private cloud or a public cloud service like Amazon EC2. The virtual cluster architecture is scaled to meet the workflow requirement, and a master node is running the SLICEE middleware. Each node of the cluster is running a bioinformatics specific Linux distribution to provide access to a wide range of bioinformatics applications. The virtual cluster has been tested on a private cloud using OpenNebula and KVM, following step is Amazon EC2 integration. All steps, starting from the cluster configuration to the workflow design and execution are performed through a web browser. The open source OBIWEE bioinformatics cloud service has been designed to allow groups with low IT support or poor computing infrastructure to analyze their own data. It also helps at facing the increasing demand for bioinformatics intensive treatments, in a context of large dissemination of sequencing technologies usages.