Drew Endy has suggested that problems such as our inability to avoid or manage biological complexity, the current tedious and unreliable construction of synthetic biological systems and the apparent spontaneous physical variation of biological system behavior can be solved using concepts such as standardization, decoupling and abstraction .
But, he suggested that evolution can be quite hard to tackle. This is because our synthetic genetic circuits become non-operational due to mutations. Sometimes, mutations can render an artificial biological system non-working in less than 5 hours. So, the problem of mutations (evolution) has been a quite difficult for synthetic biologists. It has also been suggested that we should learn from viruses, bacteria about how they manage to function normally in spite of the effect of evolution over the generations.
Here, I propose an entirely different approach to tackle evolution. According to me the concept of hierarchy and decentralization can solve the problem of abnormal behaviour of biological system due to mutations. I here talk of biological systems such as Standard Biological Parts(SBP).
I really think that SBP should create a hierarchy into the parts. Like, the parts should be defined on various levels with higher level parts employing low-level parts for its functioning. Then, we can create a redundancy of low-level parts into high-level parts so that even if a low-level part gets mutated and becomes non-functional, the high level part must keep on working normally.
A beautiful example of this approach can be TCP/IP protocol. Internet works on this protocol and thus is not dependent on any single node for its functioning. It is completey decentralised. So, even if one node (or even many nodes) come off the network, the network still functions in the normal way.
So, let us say we are making a biological equivalent of an AND gate which is composed of transistors. Then, if we design the AND gate in such a way that it is not dependent on any single transistor, we can expect the AND gate to work normally even if some of the transistors get mutated and become non-functional.
We can then scale this up. Like, we may then want to design some specific biological part e.g. an amplifier which consists of AND gates. Then, we design the amplifier in such a way that it becomes decentralized and is not dependent on any specific AND gate for its working. Now, even if a AND gate composed of many transistors gets completely un-operational due to mutations in all transistors, the amplifier will work normally.
Using this approach we can expect the increase the working life SBP quite dramatically. Also, this would enable the SBPs to work in diverse conditions.
What do you say about this approach?
 Drew Endy, “Foundations of engineering Biology”, Nature 2005.