Computation in biological processes

三个世纪前，微积分的发展为物理学的形式化奠定了基础。信息和计算科学现在处于有利地位，可以协助建立一个类似的正式框架，以严格的方式对生物过程进行推理。这里提出的这项研究的长期目标属于新兴的自然计算学科，将有助于为生物过程奠定这些正式的基础。此外，通过研究基于自然规则的理论模型，而不是人工生成的规则，我们对计算本身的本质打开了新的见解。在短期内，我们提出了两个项目：微小线虫的基因组与人类基因组的基因数量大致相同；然而，我们观察到这两种生物的复杂程度明显不同。对这种二分法的一种解释是，基因可以经历改变它们的替代处理事件，允许一个基因编码多个蛋白质。一个这样的处理事件是翻译过程中的移帧。我们已经将移码事件建模为计算，并建议研究这一模型，以揭示移码的形式属性，并开发有效的算法来检测、模拟和实验真实基因组中的移码。生化系统中自由能的最小化是驱动包括复杂大分子行为在内的化学反应的主要驱动力。到目前为止，自然计算中的大多数现有模型都没有考虑到这样的热力学因素。我们建议开发一个通用框架，将热力学考虑引入形式模型，使它们能够更紧密地反映所研究的基本物理系统，并提供一种新的计算模型。

Three centuries ago, the development of calculus laid the foundation for the formalization of physics. Information and computational science is now well-positioned to assist in the creation of an analogous formal framework for reasoning about biological processes in a rigorous way. The long-term goals of the research proposed here fall into the emerging discipline of Natural Computing and will help to lay these formal foundations for biological processes. Moreover, by studying theoretical models based on natural rules, rather than artificial human-generated rules, we open new insights into the very nature of computation itself. In the short-term, we propose two projects:The genome of the microscopic worm C. elegans has roughly the same number of genes as the human genome; yet, we observe apparently different levels of complexity in these two organisms. One explanation for this dichotomy is that genes can undergo alternative processing events that modify them, allowing a single gene to encode multiple proteins. One such processing event is that of frameshifting during translation. We have modelled frameshift events as a computation and propose to study this model both to uncover the formal properties of frameshifting as well as develop efficient algorithms for detecting, simulating and experimenting with frameshifting in real genomes.The minimization of free energy in a biochemical system is the primary force driving chemical reactions including complex macromolecular behaviours. Most existing models in Natural Computing have thus far abstracted away from such thermodynamic considerations. We propose to develop a general framework for introducing thermodynamic considerations to formal models, allowing them to reflect more closely the underlying physical system under study and providing a novel model of computation.