CCPN use of Taverna
by Rasmus Fogh r.h.fogh@bioc.cam.ac.uk, Feb 2013
WeNMR (Worldwide e-Infrastructure for NMR and structural biology) is a virtual research community project that is funded as part of the 7th EU framework. The infrastructure lies within the fields of structural biology, NMR and SAXS. Its purpose is twofold: It provides an infrastructure for sharing documentation, outreach and communications. It also offers web portals where structural biology programs can be run, supports a grid infrastructure for running associated relevant jobs and provides software that presents programs on the web portals and pipes data between them for uploading and downloading.
Our contribution to the WeNMR project is the development of a dual use Workflow Management System (WMS) as shown in Fig.1. Its two methods of use are as a:-
- means to launch, manage, and gather results generated by structural biology calculations, making use of programs present on the WeNMR portals, in effect to serve as an entry port for WeNMR and associated calculations in general.
- data tracking, management and access control system for structural biology laboratories.
Fig. 1
Taverna is being used to encapsulate and combine protocols on our system, so that the systems are available to, and can be combined with, workflow components from other sources, and to combine different protocols on our system into pipelines. The current test project is to develop a system to run four different structure calculation programs from a single job start specification and collating the results into a single combined data source.
Whether Taverna is being used in a scientific role is down to the definition of its role within the project. We are developing scientific software, which will be used to perform actual scientific research and analysis by end users – we tend to categorise our role as that of methods developers for structural biology. Whether that categorises the role as computer science, structural biology, or software engineering depends on the point of view.
The methods are established programs for calculating (mainly) macromolecular structures from Nuclear Magnetic Resonance data, and are therefore used as the primary process for obtaining results. These types of calculation require the use of several different programs, sequentially, to process them. Passing the data and results from one program to another has been a problem for many years. CCPN has long been working on data exchange standards, data conversion software and integration with different programs, and is now working explicitly on setting up pipelines. There is an ever-increasing demand for high-throughput calculations that allow the use of alternative programs and that inexperienced users are able to utilise with ease. It is to meet these demands that Taverna is being used as a tool that enables the setup of flexible, easy-to-use pipelines.
About me:
I am currently working at the University of Cambridge, Department of Biochemistry, under Professor Ernest Laue. As part of my role I am attached to the CCPN project, led by Professor Vuister (University of Leicester), formerly led by Professor Laue.
