• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>2nd international conference on bioinformatics and computational biology 2010 >A Grid-Enabled Problem Solving Environment for QTL Analysis in R
【24h】

A Grid-Enabled Problem Solving Environment for QTL Analysis in R

机译：R中QTL分析的网格化问题解决环境

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

We present a grid-enabled problem solving environment (PSE), for multidimensional QTL analysis. The concept of a computational grid has not been fully realized within this and other application fields, mainly due to the need for significant rearchitecting of problem-specific code bases. Our environment allows transparent and efficient utilization of available heterogeneous computational resources. The PSE is built around the open source statistical software system R.rnThe PSE can be used as a grid-enabled workflow system by the end-users, which is intended to be biologists with a pragmatic approach to high performance computing. Use of the Grid Job Management Framework (GJMF) offers the PSE an abstractive and reliable job management interface that supports multiple grid middleware implementations. The use of R enables a tight workflow between computationally demanding analysis and further visualization or processing steps taking place locally, or even interactively, on the user's desktop. We have developed a prototype of such a system using existing components. The architecture has been chosen to ensure transparent use of different computational resources, even from an application developer perspective, by serializing the full state for a request and sending it to a different computational node for deserialization and execution. Based on this prototype, we illustrate some example workflows and timing results from actual multidimensional searches done on experimental data, which are now possible to complete within an hour rather than over the course of several days.
机译：我们提出了一个基于网格的问题解决环境（PSE），用于多维QTL分析。计算网格的概念尚未在此和其他应用程序领域中完全实现，这主要是由于需要对问题特定代码库进行大量重新整理。我们的环境允许透明有效地利用可用的异构计算资源。 PSE围绕开源统计软件系统R.rn建立。最终用户可以将PSE用作支持网格的工作流系统，该系统旨在为生物学家提供实用的高性能计算方法。网格作业管理框架（GJMF）的使用为PSE提供了一种抽象且可靠的作业管理界面，该界面支持多种网格中间件实现。 R的使用使计算要求苛刻的分析与在用户桌面上本地甚至交互地进行的进一步可视化或处理步骤之间的工作流程紧密相关。我们已经使用现有组件开发了这种系统的原型。选择该体系结构以确保透明地使用不同的计算资源，甚至从应用程序开发人员的角度来看，也可以通过将请求的完整状态序列化并将其发送到不同的计算节点进行反序列化和执行，来确保透明地使用不同的计算资源。基于该原型，我们说明了一些示例工作流程和对实验数据进行实际多维搜索得出的计时结果，现在可以在一小时内而不是几天内完成。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号