At some point, neuroscience HAS to replace "serotonin" and other monoamine hypotheses by what they're really getting at: serotonin/dopamine/glutamine etc are all data packets for a neural network, and a better way of describing the network-level changes for depression etc is sorely needed.