infected_by to keep track of the number of nodes that can infect each node.u in initial, perform a DFS starting from u to find all nodes that can be infected by u.u and increment the count in infected_by for each node reached.v not in initial, determine if it is uniquely infected by only one node u in initial.u in initial, calculate the total number of nodes that would be uniquely infected if u were removed.u that, if removed, results in the smallest number of total infections. If there are multiple such nodes, return the one with the smallest index.