Hallmarks of disease: how tuned hierarchies of intelligent molecular neural networks (MNNs) may matter

The intracellular molecular networks are highly complex and plastic. Yet, they obey the network principles discovered by systems biology. Precise mathematical models of some networks enable one to predict how molecular properties determine the function and malfunction of the network. However, the complexity is even greater than this: due to selection in evolutionary biology, the molecular networks are not only causal to cell function, but also caused by the requirement that the cell's functions be optimal. We discuss how the combination of complexity, plasticity, and this circular causality makes the networks similar to trained artificial neural networks. Causation by purpose might thereby dominate over causation by mechanism. Comparison of challenges that may cause or cure disease with actual challenges in evolutionary, developmental, or cell biology, and assessment of the learned responses thereto, might add another interesting layer of systems biology.