Collaborative Research: Cross Species Analysis of Biological Systems Using Expression Data

美国卡内基梅隆大学（carnegie Mellon University）和匹兹堡大学（University of Pittsburgh）获得了一项资助，用于开发和实施算法和软件工具，用于分析多个物种中相同生物系统的基因表达实验。生物系统，包括在发育和癌症中起主要作用的细胞分裂过程、免疫反应和昼夜节律（控制我们一天中的行为），在许多不同的物种中都被类似地激活。事实上，理解这些系统的大部分进展都是通过研究它们在模式生物中的行为实现的。当使用表达实验（测量不同时间点和不同条件下的基因水平）来研究这些系统时，研究人员通常会识别出每个物种中表达的数百甚至数千个基因。这使得分析结果集和准确识别控制系统关键方面的基因变得困难。此外，我们正确检测这些基因之间相互作用的能力大大降低，因为许多潜在的相互作用可以用来解释观察到的表达值。这个联合计算实验项目将使我们能够识别一组核心基因；在多个物种中在序列和表达上都保守的基因。这些基因代表了生物系统的关键组成部分，对其正常维护和功能至关重要，将用于进一步的研究和初步建模。我们的方法对这些系统的预测将通过新的生物实验来验证，这些实验将反过来用于进一步完善计算模型，从而形成计算-实验循环。作为提案的一部分，我们还将开发一个用于跨物种表达分析的开源软件包。我们还将开设一门关于跨物种基因组数据分析和使用的新课程。

Cross species analysis of biological systems using expression dataCarnegie Mellon University and the University of Pittsburgh are awarded a grant to develop and implement algorithms and software tools for the analysis of gene expression experiments that study the same biological system in multiple species. Biological systems, including the cell division process which plays a major role in development and cancer, immune response and circadian rhythm (which governs our behavior over the course of a day) are similarly activated across many different species. Indeed, much of the progress in understanding these systems was achieved by studying their behavior in model organisms. When using expression experiments, which measure the level of genes at different time points and in different conditions, to study these systems, researchers often identify hundreds, or even thousands, of genes as being expressed in each of the species. This makes it hard to analyze the resulting sets and to accurately identify genes that control key aspects of the system. In addition, our ability to correctly detect interactions among these genes is greatly reduced since many potential interactions can be used to explain the observed expression values. This joint computational-experimental project will allow us to identify a core set of genes; genes which are conserved in both sequence and expression across multiple species. These genes represent key components of the biological system, are essential to its proper maintenance and function and would be used for further studies and for initial modeling. Predictions made by our method for these systems will be validated by new biological experiments which will in turn be used to further refine the computational model leading to a computational-experimental loop. As part of the proposal we will also develop an open source software package for cross species expression analysis. We will also develop and offer a new class on the analysis and use of cross species genomics data.