Study On Colon Cancer Prognostic Analysis Method Based On Multi-omics Data And Clinical Observations

Colon cancer is one of the leading causes of cancer-related death worldwide,and has caused a serious burden on people’s health and quality of life.Prognostic analysis research can support clinicians to make better decisions about the severity of the patient’s disease,and discover new disease-related prognostic factors.Therefore,it is of great importance to help oncologists better analyze the prognosis of colon cancer.The existing TNM staging system of colon cancer,which is widely used in clinical practice,uses clinical observations including the extent of the tumor,the spread to nearby lymph nodes and distant metastatic status to stage the prognosis of patients.However,large-scale clinical studies have found that the TNM staging system has an abnormal phenomenon that the survival rate of stage II patients is lower than that of stage III patients,suggesting a limited ability of personalized and precise clinical decision support.At the same time,the emergence of new technologies such as gene sequencing has brought new opportunities for precision medicine,and has already played an important role in the clinical prognostic assessment of breast cancer and other malignant tumors.However,current prognostic factors obtained solely based on genomics and other data analysis often ignored their correlation with clinical prognostic factors,making them unable to effectively complement the clinical prognostic factors,and difficult to be applied in clinical prognostic performance improving.Therefore,this paper conducted research on colon cancer prognostic method based on multi-omics data and clinical observations,multi-omics data and clinical observations were effectively combined,the current colon cancer clinical prognostic staging system were improved,potential prognostic biomarker for clinical use was identified,and a colon cancer prognostic analysis system was developed based on multi-omics data and clinical observations,to provide clinicians with a more accurate prognostic prediction tool.The main innovation points are as follows:We proposed a multi-omics integrated prognostic analysis method based on unsupervised clustering.The dimensionality of various omics data is reduced through unsupervised clustering,therefore,clinical observations and multi-omics data were effectively integrated together.The prognostic performance of colon cancer clinical prognostic staging system was improved and the prognostic model which integrated both clinical observations and multi-omics data showed significant improvement compared to current clinical prognostic methods.We proposed a method of colon cancer prognostic biomarker discovery based on integration of multi-omics data and clinical observations.The introducing of prior knowledge successfully enriched prognostic information of different gene symbols.A potential prognostic biomarker which could effectively distinguish high-risk population of colon cancer patients was discovered.In particular,it is a potential biomarker for distinguishing high-risk population of stage II colon cancer patients.We developed a colon cancer prognostic system based on multi-omics data and clinical observations.The system could provide clinicians with more accurate prognosis of colon cancer patients and tools for assisting clinical decision making.In conclusion,this thesis carried out research on the prognostic method of colon cancer based on multi-omics data and clinical observations.The prognostic performance of current clinical prognostic staging system of colon cancer was improved,potential prognostic biomarker which could distinguish high-risk colon cancer patients was discovered,and corresponding clinical decision making tools were provided for assisting clinical decision making.