基于混合神经网络的人类基因组启动子识别研究

Promoter prediction and recognition in human genome based on hybrid neural networks

提出了基于混合神经网络的人类基因组启动子识别的新方法:PromPredictor.该方法通过对转录起始位点(TSS)信息,调控区、编码区组成成分特征信息及CpG岛相关信息的综合来预测人类基因组启动子.对人类4、21、22号染色体启动子的预测结果为:敏感性达到了64.47%,特异性达到了82.2%.与其它三个算法相比,PromPredictor具有更高的敏感性和特异性.研究中所用到的数据集合及用MATLAB编写的程序代码都可以从www.whtelecom.com/Prompredictor.htm下载得到.

A new method called PromPredictor for recognizing promoter regions in the human genome is proposed. The method makes predictions of promoter by combining information about transcription start sites （TSS）, compositional features in coding and regulatory regions and CpG islands. The evaluation results on Human chromosome 21 and 22 were 64.47% in sensitivity and 82.2 % in specificity. Comparison with other three systems revealed that this method had superior sensitivity and specificity in predicting promoter regions. All datasets used in the experiments and computer codes written in MATLAB are freely available at www. whtelecom, com/Prompredictor, htm.