Kernel methods for the analysis of whole bacterial genomes

该项目解决了细菌基因组中调控区域和特定启动子识别的基本科学问题，重点关注两种具有重大社会，经济和生物恐怖意义的生物体。从机器学习的角度来看，该项目将是第一个直接针对检测监管区域的问题产生基于核的方法的项目。开发的方法将通过一个简单的基于网络的界面提供，允许世界各地的生物学家利用该方法作为一种工具，应用于逐渐扩大的细菌基因组类别，甚至真核生物。

This project addresses the fundamental scientific problem of the identification of regulatory regions and specific promoters within bacterial genomes, with a focus upon two organisms of great social, economic and bioterrism significance. From the machine learning perspective, the project will be the first to produce a kernel-based approach directly tailored to the problem of the detection of regulatory regions. The methods developed will be made available through a straightforward web-based interface, allowing biologists throughout the world to utilize the approach as a tool to be applied to a progressively widening class of bacterial genomes, and even to eukaryotes.