| Cancer is a leading cause of death worldwide,causing 8.2 million deaths in 2012.There are too many types factors that cause cancer and their amount is enormous.The paper designs analysis system based on genetic and external factors which cause cancer to improve the efficiency of research in cancer analysis.Work of this paper includes the following three aspects:First,we design analysis systems based on external factors which cause cancer.Divide analysis process of the external factors into the feature extraction and classification forecasting to design the system.The system uses a Cox proportional hazards regression model to extract the most relevant features of the composition of the feature set from properties the description of many cancer patients,Then use the method of least squares support vector machine to predict whether the deterioration degree of deterioration is still controllable.Experimental results show that the system can accurately predict the development of the cancer after a period of time caused.Secondly,we design the cancer recurrence pattern analysis system based on the phylogenetic tree.The system uses high-throughput sequencing to obtain the raw data of cancer samples,then the data preprocessing and format conversion,and finally generate phylogenetic trees.The results show that the origin of the tumor metastasis and multicentric origin of tumor recurrence patterns exist in all patients,which is consistent with the clinical pathology results.This proves that the molecular evolution analysis methods have broad application prospects in genomics research tumors.Finally,we design validation system for MP-EST which is the improved model based on the phylogenetic tree.The results show that,MP-EST in large-scale genetic data,the structural gene tree remained stable,some nodes statistical support was significantly improved,expanded the applicability of the phylogenetic tree in cancer genomics research group. |
