Potential therapeutic targets of gastric cancer explored under endogenous network modeling of clinical data.

Improvement in the survival rate of gastric cancer, a prevalent global malignancy and the leading cause of cancer-related mortality calls for more avenues in molecular therapy. This work aims to comprehend drug resistance and explore multiple-drug combinations for enhanced therapeutic treatment. An endogenous network modeling clinic data with core gastric cancer molecules, functional modules, and pathways is constructed, which is then transformed into dynamics equations for in-silicon studies. Principal component analysis, hierarchical clustering, and K-means clustering are utilized to map the attractor domains of the stochastic model to the normal and pathological phenotypes identified from the clinical data. The analyses demonstrate gastric cancer as a cluster of stable states emerging within the stochastic dynamics and elucidate the cause of resistance to anti-VEGF monotherapy in cancer treatment as the limitation of the single pathway in preventing cancer progression. The feasibility of multiple objectives of therapy targeting specified molecules and/or pathways is explored. This study verifies the rationality of the platform of endogenous network modeling, which contributes to the development of cross-functional multi-target combinations in clinical trials.