Malignant tumors are the product of the interaction between environmental and host factors. The rapid growth, invasion and metastasis of cells from cancer to tumor is an extremely complex pathological process with multifactorial effects and multi-gene involvement, accumulated through multi-stage changes. An important feature of malignant tumor cells is the instability of their genome, which causes new mutated cells to be generated continuously during the process of cell division and the emergence of mutated cell populations with different biological characteristics in tumor cells. With the completion of the Human Genome Project and the emergence of high-throughput, large-scale genetic analysis technologies, large-scale gene mutation analysis has been conducted on colorectal cancer, breast cancer, pancreatic cancer, lung adenocarcinoma, glioma, etc. It has been found that the gene mutation profiles of tumor tissues differ greatly among individuals suffering from the same type of tumor, which is the biological basis for the different clinical phenotypes of the same pathological type of tumor . From a single mutation to a simple superposition of several genetic abnormalities, it is difficult to explain the individual variability in the occurrence and development of human tumors and other complex clinical manifestations. The mutated genes in tumors are not only numerous and functionally complex, but also in a dynamic biological network. In cells, interconnected genes/proteins constitute a complex cellular network, including signaling pathways, gene regulatory networks and metabolic networks. In biological systems, no single gene performs its function independently, but must coordinate with other genes to accomplish certain biological processes as a link in the cellular network and participate in determining cellular behavior and phenotype. The mutated genes may have different effects on the network depending on their position in the cellular network, and different network structures have different tolerance to genetic mutations. In this sense, tumor is actually a molecular network disease, and it is one-sided to emphasize the role of single gene in isolation from the whole. This is the reason why some scholars have proposed that tumor research should shift from “oncogene addiction”, which focuses on the role of single oncogene/oncogene, to “network addiction”, which is based on network systems biology. Network addiction” is based on network systems biology. Malignant tumors not only involve a myriad of complex genetic mutations, but also these mutated genes are interlinked and co-exist in a delicate biological network. Through comparative biological studies of different stages of lung carcinogenesis, my laboratory found that tumor development may not be the result of simple superposition of genetic alterations in related genes, but rather a malignant transformation of cells driven by abnormal function of a network system of gene clusters involving cell growth, differentiation and other pathways. The degree of abnormality and complexity of this molecular network determines the malignant phenotype and individual differences of tumors. What stands in the way is a molecular network disease that has initiated a faulty program, and dealing with such a complex disease requires changing existing strategies and challenging inherent concepts. Research on cellular carcinogenesis and tumor prevention and treatment will remain a hot and difficult area of life science and medical research in this century. The current understanding of the nature of cancer cells is still in its infancy, and the day when people completely decode the nature of cancer cells may be the day when they understand the phenomenon of their own life.