基于DNA结合特异性权重矩阵的TAL效应物靶标预测

Prediction of TAL Effector Targets Based on Positional Weighted Matrix of DNA-binding Specificity

计算预测出TAL效应物的候选靶标有助于高效地确定TAL效应物的靶基因,进而阐明TAL效应物的生物学功能,但是目前在TAL效应物靶标计算预测方面的研究仍然很少.为了设计出预测TAL效应物靶标的有效算法,基于TAL效应物的已知靶标数据,构建了TAL效应物靶点的RVD-核苷酸关联矩阵,并将RVD-核苷酸关联矩阵归一化成RVD特异性概率矩阵,为靶标预测构造新的打分函数.实验结果表明,改进的TAL效应物靶标预测算法对大部分已知靶标预测的打分排名比已有算法更靠前,并且结合基因表达数据可以预测出TAL效应物的新的候选靶标.

Computational predictions of candidate targets for TAL effectors contribute to effectively discove-ring target genes of TAL effectors,and further elucidating their biological functions. Currently,the research on this aspect is still less. To design an effective algorithm to predict TAL effector targets,an RVD - nucleotide association matrix for TAL effector target site is constructed and converted into a RVD specificity probability matrix. A novel scoring function is developed to design an improved algorithm for targets prediction. Experimental results show that the improved algorithm ranks better than existing algorithms,and can predict novel candidate targets by using gene expression data.