BP网络与复杂疾病相关的SNPs数据分析

摘要

反向传播(BP)神经网络在基因数据分析中具有重要意义.本文首先介绍了单核甘酸多态性(single nucleotide polymorphisms,SNPs)的研究背景、分析热点及遇到的主要问题;在此基础上综述了当前常用的SNPs分析方法及其不足;随后介绍了BP神经网络在SNPs分析中的应用,如易感基因的筛选、肿瘤的诊断性分类等,并针对主要问题提出了解决方法,即用BP网络结合参数递减方法(parameter decreasing method,PDM)和聚类分析进行复杂疾病相关的SNPs数据分析.%Back-propagation neural network plays an important role and is proved significant in the analysis of genetic data. In this paper, we firstly introduced the background, hot spots and main problems on single nucleotide polymorphisms ( SJNPs ) analysis. On this basis, we reviewed the common methods in SNPs analysis and their shortcomings. Then we introduced the BP network and its applications in the analysis of SJNPs, such as susceptibility genes screening, diagnostic classification of tumors, and solutions for the main problems. Finally we put forward a strategy, which can analyze the complex diseases related to SJNPs by the BP network combining parameter decreasing method (PDM) and cluster analysis.