(1) You can define a zeta function for a graph, called the Ihara zeta function, a la Siegel's zeta function for Riemann surfaces but with prime cycles instead of prime closed geodesics
(2) Sunada observed that Ihara zeta satisfies Riemann hypothesis for a regular graph iff the graph is Ramanujan (very good approximators of regular trees)
(3) Ihara observed that for graphs of the form $\Gamma \backslash PSL_2(\mathbf{Q}_p)/PSL_2(\mathbf{Z}_p)$ for discrete subgroups $\Gamma$ of $\text{PSL}_2(\mathbf{Q}_p)$ aka $p$-adic symmetric spaces, his zeta function is the reciprocal of the Hasse-Weil z…