Terstyanszky Gabor, Kiss Tamas, Kukla Tamas, Lichtenberger Zsolt, Winter Stephen, Greenwell Pamela, McEldowney Sharron, Heindl Hans
School of Electronics and Computer Science, University of Westminster, London, UK.
Stud Health Technol Inform. 2012;175:152-61.
Molecular docking and dynamics studies are of considerable importance in a range of disciplines including molecular biology, drug design, environmental studies, psychology, etc. Using in silico tools to support or even to substitute wet laboratory work could help better focusing the laboratory experiments resulting not only in considerable saving of resources but also increasing the number of molecules and scenarios investigated. There are several software packages that support in silico modeling. However, these tools require lot of compute resources and special technical knowledge. As a result, many bio-scientists cannot use them. The paper describes a science gateway based solution which provides access to Distributed Computing Infrastructures such as clouds, desktop and service grids. This environment enables bio-scientists to execute simple molecular modeling scenarios or build more complex use-cases from existing building blocks while hiding the technical details of the infrastructure. Four scenarios have been defined and deconstructed in order to identify common building blocks supporting a large number of complex use-cases. A reference implementation for the first scenario regarding the impact on indicator species of pharmaceuticals released into water courses has been implemented on the EDGI infrastructure, demonstrating the feasibility of the approach.
分子对接和动力学研究在包括分子生物学、药物设计、环境研究、心理学等一系列学科中具有相当重要的意义。使用计算机模拟工具来支持甚至替代湿实验室工作，有助于更好地聚焦实验室实验，这不仅能大幅节省资源，还能增加所研究分子和场景的数量。有几个软件包支持计算机模拟建模。然而，这些工具需要大量计算资源和专业技术知识。因此，许多生物科学家无法使用它们。本文描述了一种基于科学网关的解决方案，该方案提供对分布式计算基础设施（如云、桌面和服务网格）的访问。这种环境使生物科学家能够执行简单的分子建模拟场景，或从现有构建模块构建更复杂的用例，同时隐藏基础设施的技术细节。为了识别支持大量复杂用例的通用构建模块，已定义并解构了四个场景。关于释放到水道中的药物对指示物种影响的第一个场景的参考实现已在EDGI基础设施上完成，证明了该方法的可行性。
