Font Size: a A A

Research On Tumor Purity Estimation Method By Considering Intra-tumor Heterogeneity

Posted on:2020-09-19Degree:MasterType:Thesis
Country:ChinaCandidate:Y GaoFull Text:PDF
GTID:2404330602950691Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Accuracy of tumor purity affect most epigenetics findings and even alter existing biological interpretations.The DNA methylation data which is one of the epigenetic mechanism is easy to obtain and widely used.So the tumor purity estimation method based on methylation data has become a research hotspot,such as InfiniumPurify and uInfiniumpurify methods.Intra-tumor heterogeneity is one of the intrinsic characteristics of malignant tumor growth.However,the above two methods ignore the influence of intratumor heterogeneity when using statistical methods to estimate tumor purity,resulting in inaccurate purity estimates and then affecting downstream related research.In view of the shortcomings in the current research,this paper analyzes the effect of intra-tumor heterogeneity on tumor purity.On this basis,a new method of estimating tumor purity,PureTumorRecovery(PTR),is proposed.The innovative results are as follows:1.In terms of the effect of intra-tumor heterogeneity on tumor purity,this paper proposes a method for categorizing tumor subpopulations in the cell layer by using mutation methylation sites.All classifications are summarized into three cases,and the classification methods are given.2.Combining the analysis of intra-tumor heterogeneity,this paper proposes a PTR method to estimate tumor purity using a regression model based on genetic algorithm.Firstly,a derivation was carried out to prove that the tumor purity and the mutation methylation site difference are in multiple linear regression relationship.Secondly,the advantages and disadvantages of each clustering algorithm and the characteristics of methylation site mutation were analyzed.The genetic algorithm clusters the methylation sites and the mean of each cluster are used as the data feature.Finally,the regression model is used to estimate the tumor purity.3.The experimental study on the real data set,compared with the existing methods,shows that the PTR method has better performance.Specifically,on the LUAD,BRCA,UCEC and COAD cancer datasets of the TCGA database,this paper compares the results of the PTR method with the InfiniumPurify method in terms of correlation coefficient,root mean square error and common statistical analysis.The result is that the PTR method has a higher accuracy conclusion than InfiniumPurify.In summary,this study systematically analyzed the intra-tumor heterogeneity,and proposed a method for estimating tumor purity using methylation data,and finally verified the feasibility and superiority of the method.The PTR method provides reference for tumor purity estimation based on the Illumina Infinium 450 k platform data,thus providing accurate data guarantee for subsequent research.
Keywords/Search Tags:Tumor purity, Methylation, Intra-tumor heterogeneity, Genetic Algorithm
PDF Full Text Request
Related items