摘要
RNA干扰(RNAinterference,RNAi)是由双链RNA(dsRNA)引起的基因沉默现象,它通过降解具有同源序列的mRNA来起作用,特殊设计的siRNA能使靶基因发生特异性沉默,起到确定基因功能或沉默致病基因从而治疗疾病的目的。在RNAi技术的应用中,通常采用的是长度为19bp,正、反义链3'端各有2个不配对碱基的双链RNA(siRNA)。但针对靶基因不同位点设计的siRNA作用效果差别很大。影响siRNA效果的因素是多方面的,这些因素的作用又是非线性的。本文在研究影响siRNA作用效果的各种因素的基础上,对已经公开发表的实验数据进行特征提取,作为BP神经网络的训练数据,并将训练好的BP神经网络用于siRNA活性预测。
RNAi is a phenomenon of gene silence induced by double-stranded RNA (dsRNA) that is homologous to target gene. RNAi can be used to identify the function of gene or knock down the targeted gene. In RNAi technology, 19 bp double-stranded short interfering RNAs with characteristic Yoverhangs are usually used. siRNA activity often changes much due to different site choosing of the targeted mRNA, the causes are many and nonlinear. In this article the authors tried to fred the motif features that affect siRNA activity and abstracted them from some published data of experiments, then they built a database, BP NN, and trained based on the database. The trained BP NN can be used to predict siRNA activity.
出处
《生物物理学报》
CAS
CSCD
北大核心
2006年第6期429-434,共6页
Acta Biophysica Sinica
基金
国家自然科学基金项目(60471003)~~