Inference-based Hypothesis-finding and its Application to Systems Biology

系统生物学是一个新兴的领域，旨在将生物有机体理解为生物系统。为了从生物数据中发现有用的信息，我们在基于逻辑的人工智能的基础上，重点研究了推理方法、归纳和推断。在归纳逻辑编程(ILP)中，推导和归纳都是用来推断假设的，并且可以用SUNE的结果发现过程来实现。然而，在现有的推理方法中，推理效率有待提高，并且必须建立假设选择机制。因此，本项目设置了以下三个子任务：(1)改进基于太阳及相关技术的假设发现方法；(2)开发基于统计和非统计方法的有效假设选择方法；(3)将这些方法应用于系统生物学问题。

Systems biology is an emergent field that aims to understand living organisms as biological systems. To find useful information from biological data, we focus on inference methods, induction and abduction, based on logic-based Artificial Intelligence. Both abduction and induction are used to infer hypotheses in inductive logic programming(ILP), and can be implemented using the consequence finding procedure SOLAR. However, in the current inference methods, efficiency of inference needs to be improved, and mechanisms for hypotheses selection are necessarily developed. Hence, the following three subtasks are set in this project:(1) Improving hypothesis finding methods based on SOLAR and related techniques;(2) Developing effective hypothesis selection methods based on both statistical and non-statistical approaches;(3) Applying those methods to problems in systems biology.