Bayesian exponential random graph modeling of whole-brain structural networks across lifespan

Descriptive neural network analyses have provided important insights into the organization of structural andfunctional networks in the human brain. However, these analyses have limitations for inter-subject orbetween-group comparisons in which network sizes and edge densities may differ, such as in studies onneurodevelopment or brain diseases. Furthermore, descriptive neural network analyses lack an appropriate genericnullmodel and a unifying framework. These issues may be solvedwith an alternative framework based on aBayesian generative modeling approach, i.e. Bayesian exponential random graph modeling (ERGM), which explainsan observed network by the joint contribution of local network structures or features (for which wechose neurobiologically meaningful constructs such as connectedness, local clustering or global efficiency). Weaimed to identify howthese local network structures (or features) are evolving across the life-span, and howsensitivethese features are to random and targeted lesions. To that aim we applied Bayesian exponential randomgraph modeling on structural networks derived from whole-brain diffusion tensor imaging-based tractographyof 382 healthy adult subjects (age range: 20.2-86.2 years), with and without lesion simulations. Networkswere successfully generated from four local network structures that resulted in excellent goodness-of-fit, i.e. measures of connectedness, local clustering, global efficiency and intrahemispheric connectivity. We foundthat local structures (i.e. connectedness, local clustering and global efficiency), which give rise to the global networktopology, were stable even after lesion simulations across the lifespan, in contrast to overall descriptive networkchanges - e.g. lower network density and higher clustering - during aging, and despite clear effects of hubdamage on network topologies. Our study demonstrates the potential of Bayesian generative modeling to characterizethe underlying network structures that drive the brain's global network topology at different developmentalstages and/or under pathological conditions.