The protein interaction network of the inherited central nervous system diseases reveals new gene candidates for molecularly unclassified myelin disorders

Mutations in numerous genes cause the inherited disorders of the white matter in the central nervous system. Interestingly, all these mutations ultimately affect myelin, even though the corresponding proteins are involved in dissimilar functions. To address this system-level issue, we assembled the myelin disease network (MDN), in which each node represents a protein (either the mutated protein or one of its interactors), while each edge linking two nodes represents the physical interaction between the two proteins. Compared with control random networks, the MDN contains more pairs of disease proteins, whose members are linked either directly or via one intermediate protein. Then, we surmised that the interactions might not only cluster proteins into functionally homogenous and distinct modules but also link the modules together. This way, even gene mutations arising in functionally distinct modules might propagate their effects to the other modules, thus accounting for a similar pathological outcome. We found, however, that concerning the function the modules are neither homogeneous nor distinct, mostly because many proteins participate in more than one biological process. Rather, our analysis defines a region of the interactome, where different processes intersect. Finally, we propose that many non-disease proteins in the network might be candidates for molecularly unclassified myelin disorders.