The intimate association between the gut microbiota (GM) and the central nervous system points to potential intervention strategies for neurological diseases. Nevertheless, there is currently no theoretical framework for selecting the window period and target bacteria for GM interventions owing to the complexity of the gut microecosystem. In this study, we constructed a complex network-based modeling approach to evaluate the topological features of the GM and infer the window period and bacterial candidates for GM interventions. We used Alzheimer's disease (AD) as an example and traced the GM dynamic changes in AD and wild-type mice at one, two, three, six, and nine months of age. The results revealed alterations of the topological features of the GM from a scale-free network into a random network during AD progression, indicating severe GM disequilibrium at the late stage of AD. Through stability and vulnerability assessments of the GM networks, we identified the third month after birth as the optimal window period for GM interventions in AD mice. Further computational simulations and robustness evaluations determined that the hub bacteria were potential candidates for GM interventions. Moreover, our GM functional analysis suggested that Lachnospiraceae UCG-001 - the hub and enriched bacterium in AD mice - was the keystone bacterium for GM interventions owing to its contributions to quinolinic acid synthesis. In conclusion, this study established a complex network-based modeling approach as a practical strategy for disease interventions from the perspective of the gut microecosystem.
Keywords: Alzheimer’s disease; complex networks; computational simulation; gut microbiota; hub bacteria; window period.