Research On Model And Algorithm Of Recognizing Cancer Driver Pathway Based On Weighted Mutation Matrix

It is generally acknowledged that driver pathway plays a decisive role in the occurrence and progress of tumors,and the identification of driver pathways has become an imperative for precision medicine or personalized medicine.Due to the inevitable sequencing error,the noise contained in single omics cancer data usually plays a negative effect on identification,and the information contained in single set of data is relatively independent.In recent years,researches combining multiple omics cancer data are more concerned than single omics cancer data.Therefore,it is important to identify driver pathways by integrating multi-omics data to improve information integrity and accuracy,and making full use of potential information from different omics data.This paper studies the problem of identifying driver pathway based on weighted mutation matrix.The main work is as follows:A weighted non-binary mutation matrix is constructed by integrating such three kinds of omics data as copy number variations,somatic mutations and gene expressions.Based on the weighted non-binary mutation matrix,a new identification model is proposed through defining new measurements of coverage and exclusivity.Then a cooperative coevolutionary algorithm CGA-MWS is put forward for solving the presented model.Both real cancer data and simulated one were used to conduct comparisons among methods Dendrix,GA,i MCMC,MOGA,PGA-MWS and CGA-MWS.Compared with the pathways identified by the other five methods,more genes,belonging to the pathway identified by the CGA-MWS method,are enriched in a known signaling pathway in most cases.Simultaneously,the high efficiency of method CGA-MWS makes it practical in realistic applications.In the CGA-MWS method,two threshold parameters ₁λand ₂λare introduced to construct weighted non-binary mutation matrix.Since the introduction of different thresholds may have a negative impact on the recognition results,and different parameter values may need to be selected for different cancers,the CGA-MWS algorithm has weak scalability.To address these issues,differential methylation analysis was performed by introducing DNA methylation data to exclude genes with high differences in gene expression data that do not have a significant impact on cancer.On this basis,the purpose of removing the two threshold parameters ₁λandλ₂ was achieved.Based on a new weighted non-binary mutation matrix,an improved identification model IMWS,is proposed with the aim of balancing the two measures in the CGA-MWS algorithm at the same value range,balanced by the harmonic mean.On this basis,a efficient cooperative coevolution algorithm ECA-IMWS is presented.Both real cancer data and simulated one were used to conduct comparisons among methods Dendrix,GA,PGA-MWS,CGA-MWS and ECA-IMWS.The results show that the genes identified by the ECA-IMWS algorithm are enriched in known signaling pathways in more cases than those identified by the other four algorithms.Simultaneously,ECA-IMWS algorithm shows high efficiency at different scales,so it has more general adaptability.In summary,in this paper,the problem of cancer-driven pathway identification is studied.Based on the cancer multi-omics data,models and algorithms for pathway identification problems was proposed,which may be useful supplemental tools for detecting cancer pathways.