The inability to profile the dynamic evolution and heterogeneity of cancer cells presents a significant barrier to effective early screening for lung cancer. By mapping pathway interactions within and beyond gene systems, one can decode their synergistic interactions underlying molecular stability and state transitions. To this end, we propose an efficient approach that derives pathway potency characterization from methylation altered sample specific networks. Applying it to lung adenocarcinoma samples uncovers a seven-state trajectory across three branching points. We show that while patient survival tends to decline within each state, it improves when the cancer transitions to a new state at branching points, which offers a more dynamic view of progression. Our approach also identifies the top feature pathways that distinguish different cancer states. Given the marked heterogeneity in pathway interplay patterns, we propose this method holds promise for decoding cancer complexity and eventually resolving critical issues in early screening.
Mi et al. (Fri,) studied this question.