基于DNA局域结构和统计物理模型识别核小体定位  

Recognizing the Nucleosome Localization Based on DNA Local Structure and Satistical Physical Model

作　　者：杜鹏宇 李前忠[1] 张璐强 陈康 DU Pengyu;LI Qianzhong;ZHANG Luqiang;CHEN Kang(School of Physical Science and Technology,Inner Mongolia University,Hohhot O1002l,China)

机构地区：[1]内蒙古大学物理科学与技术学院,呼和浩特010021

出　　处：《内蒙古大学学报（自然科学版）》2023年第1期35-42,共8页Journal of Inner Mongolia University：Natural Science Edition

基　　金：国家自然科学基金(32160216,61861035)。

摘　　要：核小体是真核生物染色质的基本单位。核小体的精确定位影响了基因组序列对结合蛋白的可及性、转录、遗传复制和重组。了解核小体在基因组的准确位置对理解真核生物的生命活动过程有重要作用。本文基于核小体的序列和结构特征及统计物理理论,用统计物理模型预测了核小体的定位。利用统计物理和信息论原理计算了酿酒酵母(S.cerevisiae)、人类(H.sapiens)、秀丽隐杆线虫(C.elegans)和黑腹果蝇(D.melanogaster)数据集中序列片段的DNA局部结构的总能量,基于核小体序列与非核小体序列的总能量差异进行分类,通过10倍交叉验证进行了性能评估。结果显示该模型具有较好的识别效能。Nucleosomes are the basic units of eukaryote chromatin.The precise localization of nucleosomes affects the accessibility of binding proteins in genomic sequences, transcription, genetic replication, and recombination.Determining the exact location of nucleosomes in genomes is important for understanding the life activity processes of eukaryotes.Therefore, based on the sequence characteristics of nucleosomes and statistical physical theory, a statistical physical model was used to predict the location of nucleosomes.Using the principles of statistical physics and information theory, we calculated the total energies of the local structures of DNAs in the data sets of S.cerevisiae,H.sapiens,C.elegans and D.melanogaster.Based on the difference in total energies between nucleosomal and non-nucleosomal sequences, the locations of nucleosomes were predicted in 10-fold cross validation.The results show that the model has good recognition performance.

关 键 词：位置概率关联矩阵 统计物理模型 核小体定位 配分函数 

分 类 号：Q61[生物学—生物物理学] O41[理学—理论物理]