An graph-based algorithm for prioritizing cancer susceptibility genes from gene fusion data

Gene fusions are widely observed in the RNA-seq data, many of which are formed by cancer susceptibility genes. The fusion gene is formed by chromosomal mutations and is an important factor in causing cancer. Studies have shown that only a small number of identified fusion genes play a role in the carcinogenesis process. Identifying those genes is important for the study and treatment of cancer. There are only few methods for measuring importance of cancer fusion genes due to the research level remaining in start stage. It is known that the importance of fusion gene can be obtained through the gene network. In this paper, the importance of cancer fusion gene based on synchronization stability theory is proposed. The algorithm evaluates the importance of nodes in the gene network based on the theory of “destructive equal importance”, and evaluates the importance of the corresponding fusion genes through the importance of gene nodes. In the process of assessing the importance of nodes, the theory of synchronization stability is introduced to relatively stabilize the gene network. The degree of damage of the nodes is calculated by using the network difference calculation method, which indicates the importance of the gene nodes. The experimental results show that the proposed algorithm has a good evaluation effect on cancer fusion gene measurement. This paper focuses on the evaluation of the importance of cancer fusion genes, and proposes a fusion gene importance evaluation algorithm, which is helpful for the identification of important fusion genes in cancer pathogenesis.