ABI Innovation: Towards Recovery of Biological Information

贝勒医学院（Baylor College of Medicine）获得了一项资助，用于开发基于连续性进化原理的网络分析新技术，以解决生物数据整合和分析的基本问题。生物学中的一个问题是，实验信息的积累速度远远快于分析速度。这些数据复杂、不完整、嘈杂，影响了它们的解释，但一个主要的困难是，因为它们来自不同的实验，所以它们是单独分析的，或者是分组分析的。这个项目的目标是将所有的生物数据连接到一个网络中，这样每一个信息都可以根据所有其他数据来理解。为了使这种方法可行，我们提出了一种新的、有效的利用进化关系的方法，将来自数百个物种的实验结果合并到一个单一的网络中，这个网络虽然庞大，但仍然很容易计算。另一个见解是，网络的邻近区域可能执行相关的生物功能。这一假设转化为精确的数学规则，在整个网络中传播信息，并尽可能地解决矛盾。在实践中，这些网络和计算工具将预测蛋白质功能的各个方面，然后在广泛的应用中进行实验测试（p53，动物生物学中的核心基因；细菌中调节压力适应的蛋白质；以及疟疾蛋白质）。计算和实验将共同评估这种新颖和通用的网络方法的价值和局限性，以发现任何生物功能的分子起源。该项目将实现许多具有更广泛社会意义的目标。首先，它将验证生物网络分析的新技术，该技术可用于分子生物学研究触及的任何领域；这包括生物技术、生物工程、纳米技术、农业、可再生能源生产和工业过程的合成生物学应用。其次，同样重要的是要注意，科学成果属于大数据分析领域，该领域涉及科学、金融、社会和国防领域。有鉴于此，第三，我们注意到，这些项目将培养博士后科学家、研究生，并为本科生提供暑期实习机会，其中包括一些来自支持少数民族教育和研究项目的学生，因此，他们以后将能够为国家利益的许多领域贡献他们在大数据分析方面的技能。

Baylor College of Medicine is awarded a grant to develop novel techniques of network analysis, rooted on the evolutionary principle of continuity, to tackle the fundamental problems of biological data integration and analysis. A problem in biology is that experimental information accumulates far faster than it can be analyzed. These data are complicated, incomplete, and noisy, slowing down their interpretation, but a major difficulty is that because they arise from different experiments they are analyzed separately, or in small groups. This project aims instead to tie all biological data into one network so that each bit of information is understood in light of all other data. To make this approach feasible we propose a new, efficient use of evolutionary relationships to coalesce experimental results arising from hundreds of species into a single network that, although massive, is still easy to compute over. Another insight is that neighboring regions of the network are likely to carry out related biological functions. This hypothesis translates into precise mathematical rules to spread information across the entire network and, as much as possible, resolve contradictions. In practice, these network and computational tools will predict various aspects of protein function that will then be tested experimentally across a broad range of applications (p53, a central gene in animal biology; in proteins that regulate stress adaptation in bacteria; and in malarial proteins). Together computation and experiments will assess the value and limitations of this novel and versatile network approach to discover the molecular origin of function in any organism. The project will fulfill many goals of broader societal significance. First, it will validate novel techniques of biological network analysis that can be used in any area touched by research in molecular biology; this includes biotechnology, bioengineering, nanotechnology, agriculture, renewable energy production, and synthetic biology applications for industrial processes. Second, it is also important to note that the scientific results fall in the area of BIG DATA analytics, which cuts across science, finance, social and national defense areas of interest. In that light, third, we note that the projects will train both postdoctoral scientists, graduate students, and offer summer internships to undergraduate students, including some from programs supporting minority education and research, who will therefore be able later to contribute their skills in Big Data analytics to many fields of national interest.