Establishment And Preliminary Application Of New Method For Detection Of Genome Copy Number Variation In Second-generation Sequencing

With the rapid development of high-throughput sequencing technology,sequencing technology has gradually developed into two poles: one is the whole genome sequencing,which is large and complete,the other is targeted region capture sequencing,which is small and precise.Compared with whole-genome sequencing technology,region capture sequencing can isolate and enrich specific regions of the genome,such as MHC and exon regions.Although it has the characteristics of good uniformity and high specificity,but due to the uneven design of probes in capture sequencing,the variation of sequencing in the region is large.compared with whole-genome sequencing data,CNVs of capturing sequencing data are more difficult to be detected.At present,a large number of copy number variation detection methods based on captured sequencing data have appeared.Although these methods are good for largescale CNV detection,the accuracy of low-frequency CNV detection is still very low.Therefore,this study did the following:(1)We have developed a novel copy number variation detection algorithm,which is characterized by: a.Using the interval in the region as the unit,detect the CNV of the interval,and then enter the individual level to detect CNV,instead of directly detecting each Individual CNV;b.Using the information of all individuals in a large sample to detect the distribution of CNV through the distribution of read depth in the interval.Assuming there is 1 CNV in an interval,the read depth in the interval obeys a threepeak mixed distribution.The new algorithm was applied to the CNV detection of PSO1000 individuals.The results showed that the gold standard detection rates of XHMM,Exome Depth and our method were 0.05,0.18 and 0.61,respectively.The gold standard detection rates of our method were 56 percentage points and 43 percentage points higher than those of the previous two methods;the false positive rates of the three methods are 0.56,0.39,and 0.23.The false positive rate of our method is 33 percentage points and 16 percentage points lower than the other two methods;The false negative rates of the three methods are 0.95,0.77,and 0.37.The false negative rate of our method is 58 percentage points and 40 percentage points lower than the previous two methods;the overlap rates of the three methods are 0.06,0.16,and 0.76.The overlap rate of this method is 70 percentage and 60 percentage higher than the previous two methods.The new algorithm improves the CNV detection method,provides technical support for CNV detection research,and provides a certain theoretical basis for the diagnosis and treatment of diseases and related disease research.(2)Using the newly developed algorithm to conduct association analysis research on psoriasis,201 CNV segments were found in the MHC area,22 of which were extremely significantly related to psoriasis,indicating that CNV played a very important role in psoriasis;Mainly concentrated near HLA-A(5),HLA-B(8),HLA-C(8),HLA-DRB1(6),HLA-DQA1(3),indicating that these Genes play an important role in psoriasis.Innovation: The first large-scale study of the mechanism of action of CNV on psoriasis.The accurate detection of CNV is very difficult.this study developed novel copy number variation detection methods,It is very challenging to use the newly method of CNV for the detection of psoriasis-related CNV.This study is the first largescale detection of CNV and the relationship between CNV and psoriasis.