| title | Taverna |
| status | 10 |
| section | Workflow-Orchestration |
| keywords | Workflow-Orchestration |
Apache Taverna [@www-apache-taverna] is a tool for developing and implementing workflows. It is an open-source software, initially called Taverna Workbench, is now a project under the Apache Incubator project. One of the key features of Taverna is that it allows developers to include many different elements together including REST web services and SOAP [@www-taverna-wiki].
When starting a scientific experiment, the main emphasis is not just the result obtained, but the procedure or method followed. Scientific workflows have thus propped up in numerous applications and experiments to reduce difficulties. Apache Taverna was introduced initially under the myGrid project. In the words of its developers,
"They provide a high-level declarative way of specifying what a particular in silico experiment modelled by a workflow is set to achieve, not how it will be executed" [@www-taverna-intro].
Experimental procedures can be automated by Taverna, integrating numerous remote and local services, including chemistry, physics, biology, music, meteorology, among many others. To sum up Taverna's many uses, it allows the experimenter to conduct complex analytics on data. All this can be achieved from a simple PC (windows or apple or Unix OS) with computational resources that can be both public and private. Taverna is mostly used by scientists who have limited access to resources and support.
Taverna's numerous features include workflow repository, service catalog, workbench, workflow components, web portals or gateways, client/user interfaces, command line, program APIs, being the more frequently used.
"Taverna performs multi-step or repetitive analysis that involves invoking several services, enables users to copy and paste results of various services, and automates processes. The workflow provenance gives a detailed trace of workflow execution such as services executed, when, which inputs were used and what outputs were produced. Currently, Tverna works with Biomart, SoapLab, SADI, R and Bioconductor" [@www-taverna-pred-analytics].
According to an article published in the Nucleic Acids Research Journal by Wolstencroft, Haines and others (2013), it is difficult to access and integrate resources that are distributed over many services, especially if they change over time and are incompatible. In this regard, the Taverna Tool Suite enables the available distribute data and analysis methods by supplementing the workflow with service discovery [@article-taverna-paper].