Workflows for metabolic flux analysis: Data integration and human interaction

Abstract

Software frameworks implementing scientific workflow applications have become ubiquitous in many research fields. The most beneficial advantages of workflow-enabled applications involve automation of routine operations and distributed computing on heterogeneous systems. Particular challenges in scientific applications include grid-scale orchestration of complex tasks with interactive workflows and data management allowing for integration of heterogeneous data sets. We present a workflow for the 13C isotope-based Metabolic Flux Analysis (13C-MFA). The core of any 13C-MFA study is the metabolic network modeling workflow. It consists of sub-tasks involving model set-up and acquisition of measurement data sets within a graphical environment, the evaluation of the model equations and, finally, the visualization of data and simulation results. Human intervention and the integration of various knowledge and data sources is crucial in each step of the modeling workflow. A scientific workflow framework is presented that serves for organization and automation of complex analysis processes involved in 13C-MFA applications. By encapsulating technical details and avoiding recurrent issues, sources for errors are minimized, the evaluation procedure for 13C labeling experiments is accelerated and, moreover, becomes documentable. © 2010 Springer-Verlag.