List Markus
, 66123.
J Integr Bioinform. 2017 Jun 10;14(2):20170016. doi: 10.1515/jib-2017-0016.
Docker virtualization allows for software tools to be executed in an isolated and controlled environment referred to as a container. In Docker containers, dependencies are provided exactly as intended by the developer and, consequently, they simplify the distribution of scientific software and foster reproducible research. The Docker paradigm is that each container encapsulates one particular software tool. However, to analyze complex biomedical data sets, it is often necessary to combine several software tools into elaborate workflows. To address this challenge, several Docker containers need to be instantiated and properly integrated, which complicates the software deployment process unnecessarily. Here, we demonstrate how an extension to Docker, Docker compose, can be used to mitigate these problems by providing a unified setup routine that deploys several tools in an integrated fashion. We demonstrate the power of this approach by example of a Docker compose setup for a drug target screening platform consisting of five integrated web applications and shared infrastructure, deployable in just two lines of codes.
Docker 虚拟化允许软件工具在一个被称为容器的隔离且可控的环境中执行。在 Docker 容器中，依赖项的提供完全符合开发者的预期，因此，它们简化了科学软件的分发，并促进了可重复研究。Docker 的模式是每个容器封装一个特定的软件工具。然而，为了分析复杂的生物医学数据集，通常需要将多个软件工具组合成复杂的工作流程。为了应对这一挑战，需要实例化并正确集成多个 Docker 容器，这不必要地使软件部署过程变得复杂。在这里，我们展示了如何通过提供一个统一的设置例程以集成方式部署多个工具，来使用 Docker 的扩展 Docker Compose 来缓解这些问题。我们通过一个由五个集成的 Web 应用程序和共享基础设施组成的药物靶点筛选平台的 Docker Compose 设置示例，展示了这种方法的强大之处，只需两行代码即可部署。
