Software Development & Core Facilities

软件开发和核心设施部分的总体目标是使MIDAS中心的研究项目中开发的新方法，算法和模型更容易为研究和公共卫生界所用。为此，我们将整合一些独立开发的工具，以创建一个连贯的计算基础设施，这是超过其方法部分的总和。特别是，我们建议开发和维护一套计算工具，用于围绕多尺度建模计算平台（MCP）进行传染病建模和分析，并针对云部署进行优化。软件开发的主要目标是： i）开发一个MCP，能够整合在不同分辨率下工作的大规模随机建模方法。我们将首先创建一个多尺度混合模型的基础上结合全球个人为基础的集合种群模型GLEAM与美国规模代理为基础的流感模型Flute。MCP架构将围绕现有的GLEAMviz客户端-服务器架构进行设计。随着为MIDAS研究界开发的模型插件开发应用程序编程接口，该系统将得到扩展。MCP将由计算密集型工具组成，因此我们计划设计和部署具有弹性功能的云计算环境， ㈡为在中心研究活动背景下开发的统计工具开发新的功能和图形用户界面，例如pomp、TranStat等的扩展。我们将开发集成的工作流程，旨在允许这些程序以自动化的方式与MCP交换数据和参数估计。 ㈢开发一个网站，作为该MIDAS中心开发的软件的共享储存库，并提供培训材料、手册和工作流程示例。我们的目标是提供最佳的可访问性，并通过使用最佳实践和公共存储库，促进MIDAS、研究社区和公共卫生机构采用所开发的软件

The overall goal of the software development and core facilities section is to make the new methodologies, algorithms, and models developed in the MIDAS Center's research projects more accessible to both the research and public health communities. To do so, we will integrate some of the independently developed tools to create a coherent computational infrastructure that is more than the sum of its methodological parts. In particular, we propose to develop and to maintain a suite of computational tools for infectious disease modeling and analysis centered around a multi-scale modeling computational platform (MCP), optimized for cloud deployment. The major aims of the software developments are: i) Develop an MCP capable of integrating large scale stochastic modeling approaches working at different scales of resolution. We will start by creating a multiscale hybrid model based on combining the global individual-based metapopulation model GLEAM with the US scale agent-based influenza model FluTE. The MCP architecture will be designed around the existing GLEAMviz client-server architecture. It will be expanded with the development of Application Programming Interfaces (APIs) for the plug-in of models developed by the MIDAS research community. The MCP will consist of computationally intensive tools, so we plan to design and to deploy a cloud computing environment with elastic capabilities, ii) Develop new features and graphical user interfaces (GUIs) for statistical tools developed in the context of the research activities of the Center such as extensions of pomp, TranStat, among others. We will develop integrated workflows aimed at allowing those programs to exchange data and parameter estimates with the MCP in an automated way. iii) Develop a website that will act as a sharing repository for the software developed in this MIDAS Center, with training materials, manuals, and workflow examples. We aim to provide the best accessibility and to foster adoption of the developed software by MIDAS, the research community, and public health agencies by using best practices and public repositories