Likelihood Ratio Research Of MeDIP-seq And MRE-seq Data

For the past few years, the next-generation sequencing technology hasrapidly developed in the field of genomics. On that basis, higher throughput RNAsequencing has a mushroom growth. As the microarray technologies, RNA-seq issimple to operate, high performing, and range of application. Meanwhile, it cansequence multiple samples and has been one of the best tools for researchingtranscript sequencing and gene expression. In addition, DNA sequencing (ChIP-seq)and methylated sequencing (MeDIP-seq) had also applied to the corresponding fieldof biology. The new data which generated by new technology is small sample size andlarge amount of a sample, more accurate and cost savings are also the characteristicsof the new data, in particular, the new data signals are discrete. So, it is a newchallenge to analyze the new data with statistical method.For the data of MeDIP-seq, there is a new method coupled with a complementaryand low cost saving, MRE-seq (methylation-sensitive restriction enzyme sequencing).Through the novel M&M (for integration of MeDIP-seq and MRE-seq) method, wecan normalizes and combine MeDIP-seq and MRE-seq data to detect differentiallymethylated regions. Using of complementary data, the method will have highlyaccuracy and repeatability.This article try to use the likelihood ratio test based on the M&M method.Likelihood ratio is a index which can reflect truth, it’s also a complex index ofsensitivity and specificity at the same time. Likelihood ratio text (LRT) provides us aobjective criterion for choosing one appropriate statistical model. In this article, weuse the likelihood ratio test based on the M&M method to analyze methylated DNAdata.