ABSTRACT
OBJECTIVE: To construct a breast cancer gene-drug network model for extracting and predicting the correlations between breast cancer-related genes and drugs. METHODS: We developed an algorithm based on the ABC principle and the association rules to obtain the correlations between the biological entities. For breast cancer, we constructed 3 different correlations (gene-gene, drug-drug and gene-drug) and used the R language to implement the associated network model. The reliability of the algorithm was verified by ROC curve. RESULTS: We identified 185 breast cancer-associated genes and 98 associations between them, 97 drugs and 170 associations between them. The breast cancer genes-drugs network contained 127 genes and 77 drugs with 384 associations between them. CONCLUSIONS: We identified a large number of different correlations between the breast cancer-related genes and drugs and close correlations between some biological entity pairs that have not yet been reported, which may provide a new strategy for experimental design for testing personalized breast cancer treatment.
