PDGPCS

'PDGPCS' is a novel algorithm designed for predicting personalized cancer driver genes, taking into account the personalized edge weight information in gene interaction networks. Cancer is a complex and heterogeneous disease, and identifying the specific driver genes for individual patients is a critical task for personalized diagnosis and treatment.
The tool follows a series of steps to accomplish this task. First, it constructs a personalized weighted gene interaction network by integrating the patient's specific gene expression data with prior knowledge of gene/protein interaction networks. This step captures the uniqueness of the patient's genetic profile and how their genes interact.
Next, the gene mutation data and pathway information are integrated into the analysis. This integration quantifies the impact of each mutated gene on various dysregulated pathways. Using a Prize-Collecting Steiner tree model, PDGPCS identifies how the mutant genes contribute to the disruption of these pathways.
The algorithm then ranks the mutant genes based on their aggregated impact scores across all dysregulated pathways, prioritizing those genes that play a more significant role in driving the personalized cancer.
Authors have evaluated on four cancer datasets from The Cancer Genome Atlas (TCGA), and the results demonstrate that PDGPCS outperforms other personalized driver gene prediction methods. Additionally, it has been verified that considering personalized edge weights in gene interaction networks can enhance the accuracy of cancer driver gene prediction.