Research Initiation Award Grant: Advancing Bioinformatics through Services Computing

Prairie View A&amp；M将参与一项名为“通过服务计算推进生物信息学”的研究启动项目。随着人类基因组的完成，生物信息学在过去十年中成为推动分子生物学研究进步的主要力量。通过微阵列和下一代测序机等先进技术提取的数据呈指数级积累。为了应对这一挑战，人们广泛寻求新的计算机技术，特别是在高性能计算和数据管理两个主要领域。服务计算是一种跨越业务服务和IT服务之间鸿沟的跨学科技术，它提供了满足这些需求的解决方案。该项目将直接解决两个问题：在服务计算环境下的来源管理和高效的元基因组分析读取映射算法。该项目的目标是开发一个基于服务计算的工作流来源模型和一个基于MapReduce的读取映射算法，以支持交互式宏基因组学分析工作流的设计和执行。基于提出的来源模型，可以捕获和管理交互式和协作的来源信息。本文提出的基于后缀索引树的读映射算法采用了“分而治之”的机制，大大减少了搜索空间，从而提高了算法的性能。通过使用MapReduce框架将算法扩展为云计算服务，进一步提高算法的性能。该项目旨在1)开发工作流溯源模型，该模型将被纳入科学的工作流系统，以支持交互式生物信息学工作流程和溯源能力；2)开发基于MapReduce的readmapping算法，支持宏基因组学分析；3)让少数民族本科生参与服务计算和生物信息学研究；4)在普雷里维尤大学开设一门新的服务计算课程。

Prairie View A&M will engage in a Research Initiation project entitled - Advancing Bioinformatics through Services Computing. With the completion of the human genome, Bioinformatics became a major force to drive the advancement of molecular biology research during the last ten years. Data extracted through advanced techniques, such as Microarray and next-generation sequence machines, are accumulated exponentially. To respond to this challenge, new computer technologies are extensively sought, especially in two major areas, high performance computing and data management. Services Computing, a cross-discipline technology that bridges the gap between business services and IT services, provides a solution to meet those requirements. This project will directly address two problems: provenance management and efficient readmapping algorithm for metagenomics analysis under a Services Computing environment. The goal of this project is to develop a Services Computing based workflow provenance model and a MapReduce based read-mapping algorithm to support the design and execution of interactive metagenomics analysis workflows. Based on the proposed provenance model, interactive and collaborative provenance information can be captured and managed. The proposed suffix index tree based read-mapping algorithm employs the "divide and conquer" mechanism, which significantly reduces the search space, and thus improves the performance of the algorithm. The performance of the algorithm will be further improved by scaling as a cloud computing service using MapReduce framework.The project seeks to 1) develop workflow provenance model which will be incorporated into a scientific workflow system to support interactive bioinformatics workflows and provenance capability; 2) develop a MapReduce based readmapping algorithm to support metagenomics analysis; 3) involve minority undergraduate students in Services Computing and Bioinformatics research; and 4) introduce a new Services Computing course at Prairie View A&M University.