Integrating systemic module inference with attract method excavates attractor modules for cyclophosphamide contributing to prostate cancer.

Objective: The complete molecular mechanism that cyclophosphamide (CPA) induces the cell death is still unknown. To further reveal the mechanism of CPA contributing to prostate cancer, we conducted analysis on gene expression profile of E-GEOD-42913 to identify attractor modules by integrating systemic module inference with attract method.

Methods: First, case and control protein-protein interaction (PPI) networks were inferred based on Spearman correlation coefficient; then clique merging algorithm was performed to explore modules in the reweighted PPI network, and these modules were compared with each other so as to select similar modules; in the following, attractor modules were identified via attract method; finally, pathway enrichment analysis of genes in attractor modules was carried out.

Results: A total of 11,535 genes were gained. A novel PPI network with 4698 nodes (20,541 interactions) was established via mapping the genes of the gene expression profile onto the original PPIs. Then, 1635 and 1487 interactions (P < 0.05) were selected to construct the destination network for CPA group and control group, respectively. Moreover, under the threshold value of overlap -threshold value of each two modules ≥ 0.5, 42 and 56 modules were separately determined for CPA group and control group. Twenty-six pairs of similar modules ([J (Sn , Tm )] ≥0.7) were gained. In the following, an attractor module which contained six nodes (15 interactions) (P < 0.05) was identified. Finally, two pathways with terms of DNA replication (P = 0.000137) and nucleotide excision repair (P = 0.024) were identified, and RFC4, POLE2 enriched in both of the pathways.

Conclusions: We predicted that during the process of chemotherapy, CPA mainly affected the pathways of DNA replication and nucleotide excision repair to induce the cancer cell's death.