CAJ | 학술논문

转录因子结合位点的识别是阐明基因转录调控机制的重要环节,准确的转录因子结合位点的预测算法将有助于人们识别转录因子的目标基因,进而研究转录因子结合位点在上游调控区中的位置对转录调控的影响.然而,目前存在的预测转录因子结合位点的算法所得结果的特异性普遍较低,因此有必要提出一种新的有效的预测转录因子结合位点的算法.本文利用JASPAR数据库上的数据,在深入分析转录因子结合位点生物学特征的基础上,构建了考虑位点保守性和伪计数的位置关联性打分方程.预测结果表明,在最佳阈值之下,该算法对转录因子结合位点的假阳率低于0.01%.
전록인자결합위점적식별시천명기인전록조공궤제적중요배절,준학적전록인자결합위점적예측산법장유조우인문식별전록인자적목표기인,진이연구전록인자결합위점재상유조공구중적위치대전록조공적영향.연이,목전존재적예측전록인자결합위점적산법소득결과적특이성보편교저,인차유필요제출일충신적유효적예측전록인자결합위점적산법.본문이용JASPAR수거고상적수거,재심입분석전록인자결합위점생물학특정적기출상,구건료고필위점보수성화위계수적위치관련성타분방정.예측결과표명,재최가역치지하,해산법대전록인자결합위점적가양솔저우0.01%.
Reliable prediction of transcription factor binding sites can help to identify the target genes of transcription factors and infer the relationship between binding sites and regulation activity of transcription factors. However, the performance of some algorithms has been unreliable with respect to poor specificity, false-positive predictions. Therefore, more efficient algorithms are required. Based on the conservation analysis of transcription factor binding site sequences, a novel position-correlation scoring function (PCSF) algorithm for predicting transcription factor binding sites is proposed by using the high quality database of transcription factor binding site sequences downloaded from JASPAR. The results indicate that under the optimal cut-offs of this algorithm, the false positive rates are lower than 0.01% for all transcription factors.