Many people seem to be overwhelmed by the current speed of technological change. But is the world really getting more complex?
If you ask my opinion as a citizen, the answer is clearly yes. There are many indications: the number of websites increases, as does the interconnectedness of mobile devices and communication networks. However as a biologist, I cannot really give you a definite answer.
You mean as to whether organisms are getting more complicated through evolution?
Yes. There is a longstanding debate among evolutionary scientists as to whether the complexity of life has increased throughout history. Common sense tells us that yes, an elephant is more complex than an Escherichia coli (E. coli). The bacteria's metabolism is in some ways simpler than ours. We are able to make products that E. coli cannot – take steroids for example. But then we are unable to produce the eight vitamins that E. coli makes for itself. So we cannot tell for sure where metabolic evolution goes.
How about life in general?
We cannot even say that for sure. More than half of all species live a parasitic life in one way or another. And we know that one general trait of parasites is that they get simpler throughout history. This is for example because they throw out the genes necessary for treating different types of nutrients, because their host provides them with a constant food source. So the jury is out as to whether complexity increases in the biological world at large.
How can you measure complexity in biology?
There is no accepted way of measuring it, as there is for example in computer science. There you have clear and universal features of computational complexity: the number of operations per time or stored data elements. In biology you can count the number of cells. So a roundworm would be more complex than a bacterium. But some argue that it is rather the number of cell types that counts. Then again, the organization of cells seems to be more important than their bare number.
Which measure do you favour?
I would definitely not push for a universal definition. Some quantities are useful operationally for specific questions or systems. For a metabolism, which is really a network of chemical reactions, more reactions certainly entail more complexity. But then this neglects the fact that some enzymes are highly regulated. Simply counting them would neglect this level of complication. We are currently working on new definitions that are especially useful to understand how evolution innovates.
This is not published yet, so I prefer not to talk about it.
Would you say that complexity is a defining feature of life?
(Hesitates) If you take complexity as the union of many different parts, then yes.
You say that innovation depends on complexity. How?
I was hoping for this question. Usually every engineer tries to avoid complications as much as he or she can. The reason is clear: a simple electronic circuit or mechanical device is much easier to produce, more efficient in maintenance, breaks down less frequently and is more accessible to understanding. But our research showed that complexity in living organisms usually allows for innovation. This is because it creates an internal robustness to change; for example when a mutation occurs in a gene or the environment of the organism changes. In this sense E. coli is very robust as it can live on dozens of different carbon sources...
...like a backup plan?
It is not so much about backup, but about alternatives. I like to compare it to the motorway system in a big city. A backup is comparable to multiple lanes on one road. But there are also multiple motorways going to many different places. When one breaks down you can find a route via other roads. I call it distributive robustness.
And where does innovation come into the equation?
This internal flexibility allows new things to arrive. It becomes easier to change one part without shutting down the whole thing or bringing it to a collapse. This ability to tinker makes robustness important for innovation. For example, we studied and classified them according to the robustness of their structure to mutations. Then we checked how many different activities each type of structure had evolved and used this number as a measure of past innovation. It became clear that robust enzymes had experienced more innovation. Nonetheless, I must say that we do not have a direct link between complexity and innovation in enzymes.
Are complexity and innovation more clearly linked in metabolism?
Yes. We know of bacteria related to E. coli that lived inside the cells of aphids for 50 million years. They have a very simple metabolism and their development has been evolutionarily stagnant and they have not been doing anything interesting for a long time. In contrast, even different strains of E. coli have very divergent metabolisms.
For understanding things you have to simplify. How do you go about studying nature’s intricacies in your research?
I use the same strategy that scientists have used since the enlightenment. First, identify the simplest possible incidence of your research problem or create a mathematically very simple model. We call it a toy model. Then, start by getting a feeling for it to inform your intuition. Starting from this intuition, you can begin to understand more complicated cases.
You wrote books for the lay public. Did you ever fear oversimplification?
I guess every scientist who is writing for the general public knows that feeling. In every simple sentence I agonise whether to add more nuance and risk losing the reader, or add less and have a problem with my conscience. In my last book, I added lots of end notes to circumvent the problem, but you cannot do everybody full justice.
Let’s talk about big data. Is it more of a blessing or a curse?
Take genomics as an example. Some scientists are frustrated by the limited knowledge that sequencing the human genome provided for medical science. But imagine the pre-2000 world without the genomic data: whenever you worked with a gene you had to find and sequence it and you did not know whether there were similar ones in the genome. This problem is now gone. So big data is absolutely a blessing. However, to make full use of it we have to shift focus from data generation to creating ways of analysing the data. And indeed, universities now hire an increasing number of professors for data analysis who are able to integrate theory and experimental data.
What recommendation would you give to biotech entrepreneurs?
Complexity is good and can be harnessed. The price that it is harder to understand is well worth paying. People should think about designing complexity into their systems so that they are more robust and ready for innovation. A practical example is robust proteins that can function in different cellular environments. In life’s evolution such proteins have brought forward more diverse abilities than their more delicate counterparts. They can also evolve new functions better in the laboratory. This is an example of how basic research can have real life implications.