CAREER: Integrated Annotation and Comparative Analysis of Metabolic Models

新的高通量测序技术已经导致了基因组数据的快速积累跨越巨大的生物多样性。这些数据现在可以用来建立模型，描述基因组变异是如何转化为表型差异的。这项研究将建立一个新的计算框架来结合所有的基因组变异并模拟由此产生的代谢表型。这个框架将是广泛适用的，因此它可以分析任何一组生物，并将包括良好的用户友好界面文档，旨在立即访问，而不需要广泛的培训。该项目将根据研究框架和数据开发一个在线游戏，用于教育学生和感兴趣的公众有关代谢活动和代谢进化的基本概念。在高级框架中开发应用程序的研究生将为本科生提供建议，这些本科生将从罗德岛大学的成功种子（SOS）项目中招募，进行暑期实习，并将开发和测试游戏界面。最后，这款教育游戏将在为期两天的科学与数学研究性学习体验工程挑战（SMILE）项目中使用，为高中生提供互动学习体验。计算工具和教育游戏系统将通过公共网络服务器和软件库发布，该项目的研究成果将通过开源文档、期刊出版物和会议报告传播。代谢网络的基因组尺度模型（GEMs）在表型预测、进化重建、功能分析和代谢工程等方面有着广泛的应用。尽管在公共文献中重建了100多个gem，但在整合基因、蛋白质、反应和代谢物的异质性信息以全面理解基因组多样性与代谢可塑性之间的关系方面仍然存在重大挑战。为了解决这一问题，本研究将重点构建一个新的计算基础设施，将GEMs模拟与泛基因组分析（即来自任何给定生物群体的基因组集合）相结合。计算基础设施将包括一个用户友好的界面，支持代谢模型的协作构建、注释和质量检查。它还将通过提供GEM比较和泛基因组分析的新算法，使全基因组代谢变异的鉴定成为可能。具体而言，该项目将重点实现三个目标：(1)实现一种新的数据格式，将异构信息集成到代谢模型的构建中；(2)构建支持代谢网络比较分析的计算工具；(3)开发一种新的算法，提高泛基因组分析中同源定位的效率，并允许不同生物之间代谢变异的比较。关于这个项目的更多信息可以在https://zhanglab.github.io/psamm/上找到。

The new high-throughput sequencing technologies have led to a rapid accumulation of genomic data across a great diversity of organisms. Those data can now be used to build models that describe how genome variation is translated into phenotypic differences. This research will build a new computational framework to combine all of the genomic variants and simulate the resulting metabolic phenotypes. This framework will be broadly applicable, so it can analyze any group of organisms, and will include good documentation for the user-friendly interface, designed to be immediately accessible instead of requiring extensive training. The project will develop an online game based on the research framework and data that will be used to educate students and interested members of the public about the basic concepts of metabolic activities and metabolic evolution. Graduate students who have developed applications in the advanced framework will advise undergraduate students, who will be recruited from the University of Rhode Island's Seed of Success (SOS) program for summer internships, and who will develop and test the game interfaces. Finally, the educational game will be used during the two-day Engineering Challenges in the Science and Math Investigative Learning Experiences (SMILE) Program to provide an interactive learning experience for high-school students. The computational tools and the educational gaming system will be released through public web servers and software repositories, and the research outcomes from this project will be disseminated through open-source documentations, journal publications, and conference presentations. The genome-scale models (GEMs) of metabolic networks have broad applications in phenotype prediction, evolutionary reconstruction, functional analysis, and metabolic engineering. Despite the reconstruction of over 100 GEMs in public literature, significant challenge remains in integrating the heterogeneous information about genes, proteins, reactions and metabolites into a comprehensive understanding of the associations between genome diversity and metabolic plasticity. To solve this problem, this research will focus on constructing a new computational infrastructure that combines the simulation of GEMs with the analysis of pan-genomes (i.e. an ensemble of genomes from any given group of organisms). The computational infrastructure will include a user-friendly interface that supports the collaborative construction, annotation, and quality checking of metabolic models. It will also enable the identification of genome-wide metabolic variations by providing new algorithms for GEM comparison and pan-genome analysis. Specifically, this project will focus on achieving three aims: (1) implement a new data format that will integrate heterogeneous information into the construction of metabolic models; (2) build a computational tool that will support the comparative analysis of metabolic networks; (3) develop a new algorithm that will improve the efficiency of ortholog mapping in pan-genome analysis and permit the comparison of metabolic variants among different organisms. More information about this project can be found at https://zhanglab.github.io/psamm/.