Establishment And Analysis Of A Knowledge Base For Cancer Risk Prediction Models

Cancer is a burgeoning health problem worldwide,and poses an increasing risk of human affliction and economic threat.In precision medicine,accurate risk assessment is a prerequisite for the implementation of risk screening and preventive treatment,and the research and development of cancer risk prediction models contribute to the early detection and treatment of cancer.A large number of studies have focused on the risk of cancer,and related bioinformatics risk prediction models have been constructed,but there is a lack of effective resource integration.Therefore,the establishment and analysis of the cancer risk prediction model knowledge base is of great significance.The research of this subject is mainly carried out in three aspects:In the first part,we used PubMed to screen the qualified articles and build a knowledge base of cancer risk prediction model CRPMKB(http://www.sysbio.org.cn/CRPMKB/).The current knowledge base contains 371 model data,including lung cancer,breast cancer,prostate cancer,ovarian cancer,colorectal cancer and cervical cancer.In the second part,this subject explored the impact of regional differences,cancer types,and model types on the accuracy of cancer risk prediction models.By means of the chi-square test,we found that the accuracy of cancer risk prediction was greatly affected by regional differences and cancer types,but it was less correlated with model types.This study suggests that developing more targeted models based on specific demographic characteristics and cancer types will further improve the accuracy of cancer risk model predictions.In the third part,this subject mainly studies the variables of cancer risk prediction model.We divided the model variables into four categories:environment,behavioral lifestyle,biological genetics and clinical examination,and found that there are differences in the distribution of various variables among different cancer types.Further,the genetic data involved in the model were summarized.Through taking 50 lung genes involved in the cancer risk prediction models as an example to perform GO and IPA enrichment analysis,the results showed that these genes were significantly enriched in p53 Signaling and Aryl Hydrocarbon Receptor Signaling which associated with cancer and specific diseases.Preemptive,Predictive,Personalized,and Participatory are the "4P" models currently favored by the medical community.We established a knowledge base of cancer risk prediction models,which provides convenience and help for patients,doctors and researchers.Through analyzing cancer risk prediction models,we provide new ideas and strategies for early diagnosis and individualized treatment of cancer with certain theoretical and practical significance.