Back to demonology

More than two years ago now, I was reading the book by the physicist Paul Davies (reference 2), with one result being the post at reference 1. Then for some reason, I was looking at it again the other week and got interested in heat pumps, pumps which seem to become fashionable from time to time – although when I took a look a year or so ago – our splendid ‘Ideal Mexico’ gas boiler is not going to last for ever, although it has served us very well – the story seemed to be that they were expensive – and it would take a while to get your money back. Then having chased the heat pump hare for a while, I came back to re-read the rest of the Davies book.

A reasonably careful reading of the first half and a reasonably careless reading of what struck me as the much more speculative second half.

Restricting ourselves to multi-cellular life, I was reminded that one of the distinguishing features of life is that all cells carry a large amount of information in their chromosomes, written out as very long strings of characters drawn from a character set with just four members – known as A, C, G and T – a lot fewer than the 256 characters used by a lot of computers. Chromosomes which in large part orchestrate the growth of a plant or an animal containing thousands of billions of cells from a single cell and which, again in large part, are carried into every cell of that plant or animal and into any children there may be, give or take a contribution from a partner.

And where the information content of chromosomes is independent of the physical substrate: that substrate will support whatever the chromosome wants to write, just as the keyboard I am typing in will support whatever it is that I want to write – be that the works of Charles Dickens or just rubbish.

Nothing that is dead, be it ever so big and complicated, can do this. Nothing that is dead can do information, can make use of information to achieve, inter alia, action at a distance. Although nothing is putting it a bit strongly: there may be some curiosities out there on the margins which I don’t know about. And modern computers, manufactured objects which might be said to have been created in our own image, certainly do do information.

One might argue that the pattern of the electrons whirling around the nucleus of some element contains information. Which it does, but that information is not independent of the physical substrate, rather it is a consequence of that physical substrate and there is no room for variation.

So one of the problems that the origin of life people have to crack, is to work out how cells came to contain chromosomes, using much the same code as they do now, perhaps as much as a billion years ago.

Davies goes on to explain that cracking the chromosomes, reading the human genome, has not turned out to be the magic bullet that some of us were expecting. Yes, we now have those very long strings of characters, and we now know a lot about the detail of how they work, say how a string of those characters is used to build a protein, but we are still a long way from understanding how it all works together to produce a adult plant or animal from an egg.

He suggests that one way into this might be to consider things at the level of information flows. Get away from the messy nuts and bolts of the chemistry of protein construction (and destruction) and things might get a bit less messy.

He starts with the cellular automata, mentioned here from time to time, for example at reference 3, and then generalises to a network. A network of nodes and edges which can be stepped forward in time in the much same way as a cellular automaton. Such a network had been used, for example, to model the regular cycling of the cells of a certain sort of yeast; a single celled organism, but a reasonably complicated one.

In this model, the nodes stand for proteins (or the genes which encode them), and nodes can be on or off. The edges stand for chemical pathways, and edges can promote or inhibit. One protein can promote or inhibit the production of another protein – or, slightly tricky, itself. Add in some voting rules which aggregate the edges coming into a node and one can step the whole network forward in time. Behaviour which might or might not settle down to a regular cycle, which might just shut down, usually depending on the starting conditions.

Davies also talks of modularity, and despite the large numbers of people working away on these very networks, I will offer a spot of modularity in a post to come.

So these networks are offered as a way of modelling aspects of life, of bringing a bit of order to an otherwise messy and chaotic world. Of modelling at a level which is low enough to capture the behaviour of interest, high enough to be comprehensible. A trick which is often difficult to pull off. But a trick which is helped along here by the fact that models with very simple rules – like those of a cellular automaton – can generate very complex behaviour. Maybe the complexity of life will turn out to have been generated with a surprisingly small tool box. Which would be good news from the evolutionary point of view; good new in the sense that a small toolbox springing out of the welter and waste (from the start of the Book of Genesis) is a lot more plausible than a large toolbox.

Contrariwise, Davies also tells us about the activities of people who look for quantum features of biology. People who do not believe in the natural order whereby physics is used to describe things on a very small scale, chemistry on the next scale up, biology on the next and various dubious disciplines like psychology and sociology after that. Generally speaking any phenomenon can be adequately described at just one of these scales. A natural order which Christoff Koch appealed to when he wrote that we had enough to do to explain the electrical behaviour of large numbers of neurons, without looking for quantum trouble. So the people in question here do exactly that, they look for quantum trouble in biology and it seems that there is quite a lot of it. So, for example, quantum physics can be invoked to better explain the detail of ATP energy transport in cells and to better explain the conversion of light energy to chemical energy during photosynthesis – both matters which are central to life as we know it.

More or less by way of an aside, Davies also suggests that if quantum computers ever make it to industrial strength, the present generation of encryption technology will be outflanked by those who can afford a quantum computer – typically government security outfits. One more example of scientists beavering away on something that is both interesting and with a whiff of money about it – without too much regard for the consequences. Although in this particular case, I would not mind if government security outfits were able to break into encrypted traffic, a nasty proportion of which is probably up to no good.

I close with the curious case of the antler of the deer. Antlers are interesting because they are large and reasonably complicated structures which grow from scratch each year, each year slightly bigger and more complicated than last year. Growth which, in many other ways, is quite like that of a human arm or leg. The present point of interest being that it seems that, given the right sort of deer, you can cut a notch on an antler one year, then next year a new antler will appear containing a tine at the very same place. Evidence that genes are not the only place where blueprints are stored. 

However, while turning Bing and Google loose turns up references 4 thru 10, a small sample from what seems to be a very busy field, with lots of people being interested in exactly how antlers grow, how they grow so fast, not least the people who like to shoot the deer carrying them, the only bit that seemed to come close to the Davies story was a paragraph on page 61 of reference 8, snipped above. But that is all not that close and I have not been able to get at the rather elderly reference 11 to which we are referred. Nevertheless, I have seen enough to see why people might be interested.

Some confusion caused by the talk of cutting notches, which resulted in my getting lots of hits about notch genes, which appear to be important in matters growth and which are named for notches in the wings of the much studied fruit fly, Drosophila melanogaster. But not obviously relevant to matters antler, even though antlers do do a lot of growing.

All in all, a book which was well worth the second look. 

References

Reference 1: https://psmv4.blogspot.com/2019/03/maxwells-demon-revisited.html. 

Reference 2: The demon in the machine – Paul Davies – 2019.

Reference 3: https://en.wikipedia.org/wiki/Cellular_automaton. 

Reference 4: https://www.msudeer.msstate.edu/growth-cycle.php. 

Reference 5: https://westernwildlifeecology.org/antlers/. From which I learn that geneticists, like software engineers, are apt to go in for whimsical names. So here we have an important bunch of genes named for the Sonic Hedgehog.

Reference 6: Deer antler – a novel model for studying organ regeneration in mammals - Chunyi Li, Haiping Zhao, Zhen Liu, ChrisMcMahon – 2015.

Reference 7: New physiological insights into the phenomena of deer antler: a unique model for skeletal tissue regeneration – Mesalie Feleke, Samuel Bennett, Jiazhi Chen, Xiaoyong Hu, Desmond Williams and Jiake Xu – 2020/2021.

Reference 8: Deer Antlers – A Model of Mammalian Appendage Regeneration: An Extensive Review -  Uwe Kierdorf, Horst Kierdor – 2010.

Reference 9: Common themes in tetrapod appendage regeneration: a cellular perspective – Bess M. Miller, Kimberly Johnson and Jessica L. Whited – 2019.

Reference 10: https://en.wikipedia.org/wiki/Notch_signaling_pathway. 

Reference 11: The role of the nervous system in the growth of antlers – Bubenik G A – 1990.