Abstract 4940: Activated signaling mechanism of glioblastoma recurrence derived from genetics, epigenetics, and genomics abnormalities
Guangxu Jin, Jun‐Feng Gao, Stephen T.C. Wong · Cancer Research · 2012
Abstract Recurrence or progession happens to certain glioblastoma (GBM) patients after given surgeries and complex treatment regimens. This implies that the pathways related to cell growth or apoptosis resistance have been activated. However, it lacks an efficient model to identify the activated pathways from a variety of data generated by the platforms on genetics, epigenetics, and genomics. Here, we develop a new computational model, called as CSB-MOO (Cancer Signaling Bridges - Multiple Objective Optimization), to address signaling mechanism associated with the recurrence or progression of GBM. We used a pathway-driven method to address the activated pathways. In the model of CSB-MOO, a new type of network elements, Cancer Signaling Bridge (CSB), was used to expand the known signaling pathways to the cancer genes (with genetic disorders) of GBM. Facilitated by CSBs, more than 60 known signaling pathways are found to be connected with the cancer genes for GBM. Then, a well-established mathematical method, Multi-Objective Optimization (MOO), was applied to identify activated pathways from multiple-type data. We used Pareto optimal to solve this MOO problem. We applied the CSB-MOO to the multiple-type data of GBM in TCGA (n=595). 15 proteins are activated in the 7 common pathways for both of the patients with and without recurrence/progression. These proteins include EGFR and JAK proteins. The patients without recurrence or progression activate another 11 activated proteins related to metabolic of amino acids (Tryptophan and Phenylalanine). Specifically, the patients with recurrence or progression activate another seventeen proteins, e.g., PIK3R1, and 2 protein complexes, i.e., [PLCG2/PLCG1] and [ITPR3/ITPR2/ITPR1], in which the SH2 domain of PIK3R1 acts an adaptor between the kinases [PLCG2/PLCG1] and the targeted proteins. In addition, these proteins facilitate the signal transduction from EGFR to downstream proteins through mediating phosphorylation. This is consistent with the fact that the recurrence/progression patients can activate the progession of GMB. The identified activated pathways classify the GBM patients with and without recurrence/progression in an accuracy of 80.89% (P < 10−4, Fisher exact test). Survival analysis indicates that the activated pathways with genetics, epigenetics, and genomics patterns accurately predict the patients’ lifespan (P = 0.0007, Chi-squared test). Acknowledgements: This work was supported by NIH grant U54CA149196 to STCW. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 4940. doi:1538-7445.AM2012-4940