中文摘要：

我们在对染色质修饰作用进行研究的过程中,发现以往的研究只关注化学修饰间的线性关系,而对于非线性关系未充分重视。为此,我们对Pokholok等（2005）人测出的酵母组蛋白甲基化修饰与乙酰化修饰数据进行了插值处理,得到了16种全基因组组蛋白修饰数据。然后对每组修饰数据在TSS位点上、下游各取1 000 bp进行对齐、平均、平滑和归一化处理。我们发现,根据酵母基因转录水平数据,可将组蛋白修饰数据分为两大类：一类是转录增强修饰群体,一类是转录抑制或转录无关修饰群体。为了揭示不同转录活性基因上修饰间的复杂关系,我们分别利用Pearson相关分析法和Reshef等（2011）人提出的MINE算法得到了以上修饰之间的相关性。最后对该结果进行分析,得到了修饰之间的一些非线性关系,同时发现同群体间修饰的关联性更强。

英文摘要：

In the process of we studied chromosome modification,we found the previous studies only focused on linear relationship between chemical modification but on the non-linear relationship.So we interpolated the data sets of the yeast histone methylation modification and acetylation that were generated by Pokholok et al.（2005）.We got 16 kinds of protein modification data sets in the whole genome.Then we processed the each of data sets to align,average and normaliz about 1 000 bp from upstream or downstream at the site of the TSS locus.According to the yeast genome transcription level data,we grouped the whole genome histone modification data sets into two categories： one modified category enhances transcription,and another modified category depresses transcription or is independent of transcription.In order to reveal the complex relationships between different transcriptional activity of genetically modified,we calculated the correlation among several modifications with two by algorithms,Pearson correlation analysis and MINE proposed by Reshef et al.（2011）.We analyzed the results and got some non-linear relationship between histone modifications.We also found that the association between histone modifications is stronger in the same modified group.