28 March 2024

Rolie Barth replies to his critics: What have pufferfishes and plasmas in common?

pufferfish (Tetraodon mbu) (wikipedia)

In this blog I will give some more clarification about my previous blog Circular causality, another secret of life?, particularly to some remarks of Gert Korthof, which I repeat in my own words.

Gerts question is this: are all components of the systems (molecules, cells, regulatory networks) that self-organize into patterns or structures not gene products? ‘All physical processes you point out are not in 'a glass of water' or any artificial laboratory environment, no, always in a cell or an organism.’

Comparing the skin of a giant pufferfish (Tetraodon mbu) and the patterns of plasmas (a glow discharge) reveals that the skin pattern as well as other biological patterns are physical of nature. The resemblance demonstrates that mutations of genes and natural selection did not make the pattern, but selected it from physical possibilities.

First, let’s start with the pufferfish [1], what a beautiful skin pattern! (see picture above). Is this a unique biological phenomena? To answer this question we turn to plasma physics, a research field in which I have worked for more than 30 years  at the FOM Instituut voor Plasmafysica. My job was to measure the temperature of the plasma electrons. In most experiments particles in plasmas are not evenly distributed but many kind of patterns can appear.
Plasma physicists of the Working Group Purwins (University of Münster, Germany) have built devices to generate low temperature plasmas [2]. A plasma is an ionized gas with free moving ions and electrons which are able to conduct electrical current. Making such plasma between two parallel plates (electrodes) the physicists found all kind of patterns. The patterns show up at the two plates due to the plasma-surface interaction and are made visible using one transparent electrode. A digital camera is applied to record the images. The optical pattern corresponds to the current density pattern in the plasma.

At certain values of the AC voltage a pattern was found that is quite similar to that of a pufferfish. In the figure below tree patterns are shown: left – a pufferfish skin [1], middle – a pattern of the plasma surface interaction [2] and right – a simulation using the Gray-Scott model [3].
Comparing a part of the pufferfish skin (left) and a pattern of the plasma surface interaction show a remarkable resemblance. Indeed, they are not identical but a great similarity cannot be denied.

The third image (right panel) shows the result of simulations using some kind of Turing model, the so called Gray-Scott model. All these modeled processes belong to the large family of reaction diffusion processes for two ore more species of particles, which can be: electrons, ions, chemicals, morphogens, cells and also larger entities. Analyzing these reaction diffusion processes can partly be done analytical, but to determine patterns requires computer simulations.

So, what have pufferfishes and plasmas in common? In both cases reaction diffusion processes created similar patterns that correspond more or less to mathematical simulations of these phenomena. It is obvious that the patterns produced by the interaction of the plasma with the transparent plate (electrode) are the result of physical laws of nature without the need of any external information (like genes in the biological situation). This demonstrates that biological patterns like that of the pufferfish were not invented but discovered by evolutionary processes. These patterns result from physical mechanisms. Of course, in the biological pattern formation genes play important roles:
1. Most morphogens are proteins, build by genetic information.
2. Feedback loops of these systems consist of genetic networks and physical mechanisms, like diffusion.
3. Genes save information about how to reuse the physical process during next generations.
4. Contrary to purely physical processes, the biological variant of these processes can be modified by gene mutations or even by epigenetic processes.

Finally, the resemblance of pufferfish and plasma patterns also show some important features of the position information model of Lewis Wolpert, described by Marleen in reply to my previous blog. She wrote: “The patterning is well described, but can also be described by a simple gradient of morphogens that have to pass a threshold value”, where (in my own words) every threshold is genetically coded. This model is suitable to understand the creation of rather simple, linear patterns like the segments of insects, the swivels of backbones, tails of dino’s and the like. But simulating a pufferfish pattern with this model would require a lot more genetic information than describing it by a Turing model. The model of Wolpert requires two genetically coded threshold values for each stripe … In the Turing model the number of stripes and their separation distance is a matter of tuning two of the system parameters and not of adding new parameters.

In conclusion: the fact that the
pufferfish skin pattern and plasma patterns show such close similarities clearly demonstrates that these patterns are generated by physical mechanisms. Of course, genes are not unimportant because they produce the morphogens and are part of the feedback loops necessary for the pattern generating processes. Under these patters lie the physical rules and mechanisms for the complex interaction of a large number of particles. Generally spoken, these patterns are formed by the collective behavior of interacting particles which are part of non-linear dynamical systems, like physical plasmas and biological tissue.

 

References

  1. Pufferfish image: Wikipedia
  2. Plasm pattern and plasma experimental setup: Juan Pablo Trelles, ‘Pattern formation and self-organization in plasmas interacting with surfaces’, 2016, J. Phys. D: Appl. Phys. 49, 393002, Figure 3. See also: H. -G. Purwins, 2011, IEEE Trans. Plasma Sci. 39-11, 2112.  
  3. Simulated pattern: simulation for a Gray-Scott model, k = 0.61 and F = 0.42. Take look at this website with a extended simulator of such patterns: https://www.mrob.com/pub/comp/xmorphia/index.html.
     

hier nog een geweldige foto van de pufferfish !

43 comments:

  1. Patronen veroorzaakt door reactie diffusie vergelijkingen zijn onder invloed van allerlei genen die de snelheden van de diffunderende stoffen of de tijd van begin van de aanmaak van die stoffen bepalen,.
    Klassiek zijn de zebrastrepen en de patronen op schelpen. https://link.springer.com/article/10.1007/BF00289234
    H. Meinhardt 2009,The Algorithmic Beauty of Sea Shells

    ReplyDelete
    Replies
    1. Thank you, indeed many more examples can be given. Meinhardt has written several books about biological pattern formation.
      This one is free available: 1982 - Models of Biological pattern formation - Hans Meinhardt.

      Delete
  2. Rolie, thanks for your interesting contribution. You wrote: "plasma physics, a research field in which I have worked for more than 30 years at the FOM" ... Impressive! I don't understand a iota of plasma physics! How hot is this plasma?
    You wrote:
    "So, what have pufferfishes and plasmas in common? In both cases reaction diffusion processes created similar patterns that correspond more or less to mathematical simulations of these phenomena."
    and:
    "These [biological] patterns result from physical mechanisms. "
    and:
    "Of course, in the biological pattern formation genes play important roles:"
    and then you give 4 ways in which genes are involved in these patterns. (I agree).
    But, if genes are always involved, doesn't that mean that it is NOT a pure physical processes?
    It is diffusion in an environment with cells, membranes, cell organelles, etc. That is quite unlike pure physical diffusion. The biological molecules that are 'diffusing': how big or small are they? can they freely move without bumping into other molecules? are they really passively moving? or maybe actively transported? can they cross cell membranes? actively or passively? etc. etc. etc.
    ===
    Previous commenter linked to:
    A. Gierer & H. Meinhardt A theory of biological pattern formation
    Published: December 1972!
    that's long ago!

    ReplyDelete
  3. Gert, the temperature of the plasmas of which the pattern are shown, is about 10.000 Kelvin (~ 1 eV). The plasmas temperatures I measured were much higher, up to 1000 eV.

    Your question: "But, if genes are always involved, doesn't that mean that it is NOT a pure physical processes?"
    In principle I agree, the pufferfish skin pattern is the result of the interaction between genes, morphogens and physical mechanisms. And thus not a PURE physical phenomenon. Morphogens are molecules, most are proteins with a size of, 5 to 10 nm. They move freely through the space between cells and regularly collide with other molecules and proteins, due to their thermal motion.

    In general, pattern formation arises from specific interactions between a large ensemble of ‘particles’ and differences in diffusion velocities of these particles, together they constitute a dynamical system. ‘Specific interactions’, because they include positive and negative feedback loops (see previous guest blog). In physics these kind of systems are known as reaction-diffusion systems, which can be modelled with mathematical methods developed in the field of physics.
    You find such systems in pure physical systems, like the plasmas described above. But also in biological systems, like the skin tissue of a pufferfish. So the ‘particles’ can be very different: electrons and ions, but also morphogens, cells, neurons, cars, populations of organisms and so on. The created patterns appear to be very similar and therefore I call them universal patterns. No matter what kind of ‘particles’, these systems self-organize into patterns when they have similar kind of interactions and feedback loops.
    Conclusion: to be correct we should speak about plasma-physical, biophysical, socio-physical, neuro-physical, eco-physical and so on.

    ReplyDelete
  4. One thing needs to be emphasized: in the biophysical situation genes are needed, but in the plasma physical system no extra information is needed. Thus genes are important to create such patterns but in general not necessary.

    ReplyDelete
    Replies
    1. You wrote:
      A. Gierer & H. Meinhardt A theory of biological pattern formation, Published: December 1972!
      that's long ago!

      Indeed, this stuff was allready known long time ago. Faster computers and development of clever mathematical methods gave the research in this field a boost.
      Now research is multidisiciplinar (as you said above) and bundled in so called Systems Biology.

      Delete
  5. Rolie, you wrote: "They move freely through the space between cells and regularly collide with other molecules and proteins, due to their thermal motion."
    However, morphogens are produced in the cytoplasm of the cell: so if they have to 'freely move', the morphogens first have to get transported through the cell membrane. There are no (ordinary) holes in the cell membrane: it would mean it is a leaky membrane! So, special biological mechanisms (proteins) must exist to do the transport. So, the transport is under biological/genetic/cellular control. That shows that a simple physical diffusion model cannot be applied.

    You think like a physicist about diffusion!
    You have to think like a biologist!
    If you place a tea bag in a cup of hot water, you see diffusion in all directions and the end result is a HOMOGENEOUS solution! NOT a gradient!!
    In embryos and organisms things like for example morphogen degradation takes place. These processes do not take place in physics of diffusion where molecules stay the same in the whole process.

    In biology there is active transport not present in a pure physical environment:
    "The process of endocytosis is imperative to the two central themes in gradient formation: active transport facilitating long-range signaling and degradation of morphogen to sustain gradient shape. ..."
    from: https://pubmed.ncbi.nlm.nih.gov/21654212/

    "Morphogen gradients are crucial for the development of organisms. The biochemical properties of many morphogens prevent their extracellular free diffusion, indicating the need of an active mechanism for transport."
    https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1009245

    And if you have free time these days here is an overview:
    Patrick Müller et al (2013) Morphogen transport, Development, 140 (8): 1621–1638.

    conclusion: in biology things are more complicated then in physics.
    There are several competing hypotheses about morphogen transport. The question has not yet been solved...!

    ReplyDelete
  6. Gert and Rolie,

    It is now clearer what Rolie meant by a physical mechanism in patterning in organisms. Thank you very much for the clear explanation. The images of the pufferfish skin next to the plasma surface are very interesting. There must be some law in these patterns in living organisms and in dead matter. But this similarity does not mean that there is a pure physical basis for it in the pufferfish. That's not how it works in biology.

    Perhaps it has already been figured out what the mechanism is that both ‘systems’ have in common? I would like to know more about it. Perhaps both biologists and physicists need to learn more from each other

    However I do completely agree with Gert. It is very unlikely that such a pattern is created solely by diffusion.
    Thanks Gert for the reference to the article in Development.

    ReplyDelete
  7. Gert,
    Ik ben nog even langs gegaan bij je blog over Ionica Smeets en haar ananas. Dat lijkt me ook patterning.
    Qua verwondering lijkt me dat net zo hevig als de zebra.
    Maar is er ook verschil in de bedoeling?

    Kan het bij het ene schepsel functioneel zijn, terwijl het bij het andere mooiigheid is?
    En zit het in het bouwplan of kan het ook toeval zijn (ooit kan er een zebra met een kruiswoordpuzzel op zijn lijf de hoek omkomen, à la de black swan van Popper).

    ReplyDelete
  8. Hoi Leonardo, ja inderdaad Zomergast Ionica Smeets, Wilfried de Jong, Fibonacci, Ananas, God was een topper. Ananas patronen moeten mechanisch verklaard worden, en hebben verder geen 'bedoeling'. Naar de zebrastrepen is serieus onderzoek gedaan: een oudere theorie zegt: camouflage (maar misschien valt hij juist meer op?), en andere theorie zegt warmteregulatie en de meest waarschijnlijke is dat steekvliegen niet landen op oppervlaktes met dunne zwart-witte strepen. Hij raakt in de war als hij dichtbij komt en ziet niet meer dat hij met een dier te maken heeft. Dit is onderzocht door paarden dekens te geven met verschillende zwart-wit patronen inclusief schaakbord patronen!) en te kijken wat vliegen doen! Kijk: dat is nu echte experimentele wetenschap!
    Bron: We Now Know Why Zebra's Stripes Are So Effective 26 Feb 2023.

    ReplyDelete
  9. Gert and Marleen, thanks, we are getter closer to a better understanding.
    First some remarks. Gert you wrote: “You think like a physicist about diffusion! You have to think like a biologist!’ Regarding the latter: I tried to learn something about biology the last 10 years, as you can see in my book. But the point in this discussion is that scientists studying life processes need to think both as a biologist as well as a physicist, both with knowledge about the behavior of non-linear dynamical systems. A job to be done in team work.
    Thus, Gert, it is now time for you to learn thinking as a physicist – a little bit 😊

    Unfortunately, you both didn’t fully get my point about diffusion.
    Gert you wrote: ‘If you place a tea bag in a cup of hot water, you see diffusion in all directions and the end result is a HOMOGENEOUS solution! NOT a gradient!!’
    Of course you are right, I said this several times before. You find it explicit in my book (p. 198): a mixture of red and blue printer ink in a glass of water doesn’t result into a pattern of stripes but into a homogeneous purple fluid.
    And Marleen you wrote: ‘However I do completely agree with Gert. It is very unlikely that such a pattern is created solely by diffusion.’ Continuously, I spoke about reaction-diffusion processes, so indeed not solely diffusion.

    Rethinking my blog and comments, I have to admit I forgot something important written in my book: biological pattern formation has to do with three categories of lawful operations (p. 278, 336, 352, 355):
    1. Biological and chemical rules for the production and transport of morphogens and the interaction of them with genes,
    2. Physical laws of diffusion and mechanical interactions,
    3. Lawful behavior of an ensemble of whatever kind of particles subjected to positive and negative feedback mechanisms; in short: the logical rules for systems with circular causality.

    Referring to Marleen, item 3 is what you mentioned saying: we have to look for what “both ‘systems’ have in common” [RB both systems is referring to pufferfish and plasma].
    The logical rules for non-linear dynamical systems with circular causality can be explored with mathematical models as the whole family of Turing models, of which the Gray -Scott model is just one. These explorations show that specific interactions and diffusion may result in patterns which are universal, that is: they self-organize in these patterns independent of the kind of particles involved.
    And of course Gert, gradients may be (but not necessarily) be part of such systems. See for example the BSW-model describing the embryonic growth of finger structure, research of the James Sharpe and his team.
    And, of course morphogens are made in cells and subsequently transported to the extracellular space via endocytosis, no doubt about that.

    Finally, one more clarification: the interactions between particles are local, at some position in the pattern, while diffusion has a global effect. That’s why Eshel_Ben-Jacob spoke about “the diffusion field”. Pattern formation arise from particle-field (or wave) interactions. In that respect plasmas are more complex than pufferfish skin patterns, because in a plasma there are two more global factors: electrical and magnetic fields.

    References.
    Eshel_Ben-Jacob, ‘From snowflake formation to growth of bacterial colonies. Part I. Diffusive patterning in azoic systems’, Contemporary Physics 34 (5), 1994, p. 247-273.
    James Sharp see: Jelena Raspopovic et al., ‘Digit patterning is controlled by a Bmp-Sox9-Wnt Turing network modulated by morphogen gradients’, Science 345, 2014, p. 566-570.

    ReplyDelete
    Replies
    1. A small note: the biological process to transport a morphogen out of the cell is of course exocytosis. Endocytosis is transport of molecules from the extracellular space into the cell. See: https://nl.wikipedia.org/wiki/Exocytose

      Delete
  10. Rolie,


    Thank you very much, very clear again. I still have some problems. I don’t know what plasma really is and what the figure from your reference article really shows.
    I found an article from 2022, but I lack the scientific background to understand it. It is not clear to me exactly what Turing found and how it is still used and applied today. I should study it better, so I am just a layman in this respect.
    https://www.frontiersin.org/articles/10.3389/fcell.2022.871950/full

    In this article I immediately have a problem with the definition of pattern. Figure 1 speaks of Turing patterns, but the detailed photos show shapes that, in my opinion, are not really patterns. The stripes of zebras and puffer fishes are clearly patterns. So that makes it difficult to understand the rest of the article.

    It says in fig.2C

    The ornate boxfish, Aracana ornata, native to waters off South Australia; female (left) and male (right) showing mixed stripe and spotted patterns characteristic of boxfishes, which often vary between the sexes despite, presumably, a common underlying mechanism. in 2022.

    The underlying mechanism is not yet known! I assume that even though Rolie says that plasma patterning is more complex, the mechanism of it is known. It must now be examined to what extent this mechanism corresponds to the biological mechanism. Only then can it be seen to what extent the pattern formation in the fish is physical, but above all, how it exactly works. Because that is still not clear. So far we have only concluded that there must be somehow a similarity between biological and physical pattern formation.

    How can you say that: In that respect plasmas are more complex than pufferfish skin patterns, because in a plasma there are two more global factors: electrical and magnetic fields. Even if you use the words 'in that respect' . A pattern in the skin of a fish involves the transcription of proteins, post-translational changes and the exocytosis of morphogens, before they or other particles can participate in diffusion. That seems quite complex to me, more complex than plasma patterning.

    This is very difficult to understand: you write: These explorations show that specific interactions and diffusion may result in patterns which are universal, that is: they self-organize in these patterns independent of the kind of particles involved. What you write here is clear, but I can hardly believe that this self-organization of particles takes place independently of the type of particle. This would mean that the patterns of, for example, the pufferfish or the boxfish, arise from self-organization of the morphogens and that this is independent of the type of morphogen or particle. So why are the patterns different in the pufferfish and the boxfish. What makes these have such different patterns. What makes the skin self-organize once in this way and the other time in that way. If it depends on self-organization the pattern should always be the same, since the physical laws are the same.


    It is worth mentioning that Cees Dekker is also interested in patterns in E.coli:

    https://www.frontiersin.org/articles/10.3389/fphy.2022.930811/full#h6

    ReplyDelete
  11. Marleen thanks for the link to:
    Patterning, From Conifers to Consciousness: Turing’s Theory and Order From Fluctuations.
    I found an important warning in the article: "For theorists, there have been disappointments along the way, in that patterns that appeared to match theoretical prediction were shown to arise by other mechanisms."
    Indeed. I did worry about the ease by which Rolie claimed things as What have pufferfishes and plasmas in common.
    My conclusion: if patterns of pufferfishes and pasmas look similar, this proves nothing beyond they look similar.
    Leonardo already pointed to Popper, which triggers the ultimate Popper question: are all those theories of pattern formation falsifiable? It seems to me those theories can easily be fitted to the data by fine-tuning the parameters. I played a little with Gray-Scott model: it is indeed fun. Those theories can never fail. Change the model and the parameters a little and everything can be produced. Also with reaction-diffusion models: change the parameters, add some powerful 'attractors' and they predict whatever you want. They are always true. Have there been serious attempts to falsify those theories? The only way is to do experiments. And that is the hard thing to do. Very hard.

    ReplyDelete
  12. Marleen, thanks for the interesting review paper.
    “Patterning, From Conifers to Consciousness: Turing’s Theory and Order From Fluctuations.”

    The author gives a nice review of the subsequent developments in application of Turing’s ideas about pattern formation. But the title and the abstract show a serious mistake: Order From Fluctuations. Fluctuations, as shown in my first blog, trigger the beginning of the pattern formation, but fluctuations are not the creatives source. It is circular causality and in particular the positive feedback loop of such systems that may lead to patterns.

    Referring to the subscript fig 2C:
    The ornate boxfish, … native to waters off South Australia; female (left) and male (right) showing mixed stripe and spotted patterns characteristic of boxfishes, which often vary between the sexes despite, presumably, a common underlying mechanism.

    Here, “presumably” does not refer to the question whether we know the kind of underlying mechanism but asks whether both male and female pattern are the result of the same mechanism – I add: with some small variations in the interaction parameters of morphogens with genes.

    More important, it seems to me that you and Gert both think that the great resemblance between pufferfish and plasma patterns is nothing more than a look-alike similarity. I wonder how you can make such judgement … Do you really think that a whole university department (Systems Biology at Harvard, or Mathematical Biology at Utrecht Univ.) would invest money and time when it is all a look-alike?
    Is there any experimental evidence? Yes, take a look at these publications:
    Jelena Raspopovic et al., ‘Digit patterning is controlled by a Bmp-Sox9-Wnt Turing network modulated by morphogen gradients’, Science 345, 2014, p. 566-570.

    Raspopovic and colleagues modelled the growth of mouse fingers and compared model and real world while blocking different morphogens. They found similar distortions of the finger pattern. More examples can be given:

    In 2006, Stefanie Sick et al. (2006) describe two transcription factors WNT and Dkk1, that are involved in the development of mouse hair follicles. The protein WNT fulfills the role of activator, while Dkk1 has the function of an inhibitor. A Turing model for this hair growth predicts that a moderate overproduction of the inhibitor Dkk1 will lead to an increase in the distance between hair follicles and experimentally work confirms that. And the other way around, the model shows that a moderate increase in the production of the activator WNT will lead to a higher density of the hair implant and this also appears to be the case in experiments with mice.
    Stefanie Sick et al., ‘WNT and DKK determine hair follicle spacing through a reaction-diffu¬sion mechanism’, Science 314, 2006, p. 1447-1450.

    Shigeru Kondo et al. (2010) researched the mechanism behind the growth of stripes in zebrafish. They also found striking similarities between model and reality. The studied species zebrafish has approximately five horizontal stripes on the body. In an experiment the middle part of two of those stripes were burned away with a laser. After two weeks, the nearby stripe begins to deform and shifts into the open space between the two interrupted (and damaged) lines. This continues until the shifting line completely lies between the two interrupted stripes, curved in a bell shape. Simulation with a Turing model for the growth of the stripes demonstrates exactly the same shifted bell-shaped pattern (Yamaguchi, 2007).
    Motoomi Yamaguchi, Eiichi Yoshimoto and Shigeru Kondo, ‘Pattern regulation in the stripe of zebrafish suggests an underlying dynamic and autonomous mechanism’, PNAS 104 (12), 2007, p. 4790-4793.

    In conclusion, Gert, Marleen and Leonardo: yes, these reaction-diffusion models are falsifiable, but there is a long, difficult wat to go.

    ReplyDelete
  13. And finally, I was a little bit joking when I wrote: ‘In that respect plasmas are more complex than pufferfish skin patterns’, of course Marleen you are right biological process are very complex. But in every joke is some truth: the mathematical modelling of the plasma-wall interactions are very complex compared to a basic Turing-model.
    The point is that most biologists have no idea about the mathematics behind plasma physics (and I myself am a layman in this matters).

    By the way, Wikipedia gives a nice article about plasmas: https://en.wikipedia.org/wiki/Plasma_(physics)

    To be continued

    ReplyDelete
  14. Gert, one more remark: most of what I wrote above (including the experimental verifications) can be found in the book of Philip Ball "How life works" (p. 307-328) as well as in my own book (De kosmos en het leven, een Meesterwerk, p. 264-271).

    ReplyDelete
  15. Rolie and Gert,

    The genuine question is, what does the article Digit patterning is controlled by a Bmp-Sox9-Wnt Turing network modulated by morphogen gradients have to do with plasma?

    I don't think we doubt Turing's model and the reaction-diffusion system of morphogens. I rather think that we (I) cannot understand specifically how plasma can have the same pattern as that of a pufferfish, or vice versa.
    It seems as if Gert thinks that it is a matter of the right input after which the plasma forms certain patterns, some of which 'coincidentally' look the same as those of the pufferfish, but I am not sure that is what he means.

    The rest of the articles Rolie cites are all about growth factors/morphogens. That's not the problem. That concerns biology and we have no doubt about that. The hair follicles are also very interesting, and that the Turing model correctly predicts the result of these experiments. However, this is not about plasma. And it is not about Turing and plasma together. It is about computer models and experiments whose results match.

    The abstract of the article about the stripes of zebrafish closes with: Because the zebrafish is the only striped animal for which detailed molecular genetic studies have been conducted, our finding will facilitate the identification of the molecular and cellular mechanisms that underlie skin pattern formation
    The mechanisms were not yet known at the time (in zebrafish: 2007). I'm curious to see how things stand now.

    I had seen the article about plasma in the English Wikipedia, but it did not really provide clarity about the patterns. I have now started reading Turing 1952, which will probably prove too difficult.


    ReplyDelete
  16. Dit lijkt ver weg maar ik vind het toch wel dichtbij. Ik wilde iets zeggen over Turing en modellen én mathematica dat niet stuurt maar gestuurd wordt door verschijningsvormen in de natuur (kom ik ook nog wel mee) maar dat werd even op de plaats rust gezet. Want de Volkskrant kopt Narcisten worden niet geboren, maar gemaakt.

    We weten dat dieren allerlei “psychische” trekken laten zien die het dier Mens ook vertoont – het schuldgevoel bij apen bijvoorbeeld. Ik ben een geharnast voorstander van het mind-body dualism – het zou ook body-mind moeten zijn, body als principaal, zonder de body is er niets. Ik ben er van overtuigd dat de materie, inwendig of uitwendig én de fysische processen daarvan in het lichaam de inhoud van de hersenen bepalen.
    Wel, als vissen narcistisch kunnen / zouden mogen zijn, dan toch in ieder geval de puffervis – what a beautiful skin pattern! zegt Rolie, met uitroepteken.

    Dus, hoe zit het dan met die blueprint? Het mechanistische wereldbeeld. Het materialisme.
    Van een man met een bochel en een kromme neus kun je toch geen narcist maken?
    En zonder blueprint geen modellen!

    Het enigma Leven – daar begon het toch allemaal mee – blijft een enigma.

    NB De eerste vraag had natuurlijk moeten zijn: zijn de dieren, evenals de mens, eigenlijk ook wel zo geobsedeerd door patronen?

    ReplyDelete
  17. Rolie, Gert and Leonardo,

    This is an extremely extensive subject that Rolie has brought up.
    Turing's 1952 paper is quite mathematical. It seems he doesn't yet know how genes work, which is understandable. On page 40 he writes
    The effect of the genes is presumably achieved almost entirely by catalysis. They are certainly not permanently used up in the reactions.
    He seems to assume that genes participate in a reaction between or from morphogens.
    There is a lot to be found on the internet about Turing pattern and Turing instability. A very nice article from 2017 in which the Meinhardt & Gierer model is elaborated 'better' or 'more biologically' is the following:

    An updated kernel-based Turing model for studying the mechanisms of biological pattern formation

    It's a very interesting topic. Thank you very much for these blogs.

    I once heard about the slightly inhomogeneous structure of the universe just after the Big Bang that led to the coming together of matter into the forms we know today. Would that be correct?

    ReplyDelete
  18. Rolie, when I pointed to the fact that the belly of the pufferfish has no pattern, you replied that certain mathematical simulations can easily produce the featureless belly of the pufferfish. This got me thinking: if those models can explain the presence AND the absence of patterns, then they can explain anything! Do those models forbid any pattern? Are there patterns in the natural world that canNOT be explained by theoretical models? I am inclined to think: if they explain anything, they explain nothing. Again: falsifiability! Ignoring falsifiability is a scientific mortal sin.
    Yesterday I quoted from Marleen's article:
    "For theorists, there have been disappointments along the way, in that patterns that appeared to match theoretical prediction were shown to arise by other mechanisms."
    That casts doubt on the usefulness of mathematical models. One should at least be very careful. You cannot take them to be true simply because they predict patterns seen in nature.
    Yesterday I read an article in Nature about theoretical physicist/cosmologist Claudia de Rham who criticized Einstein and had really important and wise things to say about science:
    "Every day, I try out an idea and it fails. And there’s something beautiful in failing, and falling. The book is about gravity, but it is also about embracing this falling, because it’s how we get better — it’s how we understand the world. ... So the theory itself is telling you that you shouldn’t trust it any more. And that is not something to be ashamed of. It is an opportunity to learn something more."
    Precisely! That's why it is very wise to study the reasons why mathematical models fail and learn the lessons from it. In order to become a better scientist and have better theories.

    ReplyDelete
  19. Marleen thanks, you found a very interesting paper that gives (to me) new insights in the modelling of biological patterns and structures. I will read it with pleasure.

    Gert, referring to your remark: "if those models can explain the presence AND the absence of patterns ... etc." Here I lose you, forget Turing patterns for a moment and think of snowflakes. Observations and models both show that such patterns ONLY arise under certain circumstances. In reality and in models there are much more circumstances where no snowflakes are formed, fortunately it is like that 😊. Do you agree? The same is valid for all scientific models.

    You continue: “I am inclined to think: if they explain anything, they explain nothing. Again: falsifiability! Ignoring falsifiability is a scientific mortal sin.” Consider this in the light of my words above. A model that could not predict how patterns WILL or WILL NOT be formed, depending on circumstances and other system parameters would be rubbish. So what you call a weakness of Turing models is a normal property of models and real situations.
    In my eyes your statement makes scientifically no any sense.

    And of course, mathematical models don’t relate 1-to-1 with reality but that doesn’t make them wrong by definition. No, models have to be verified with observations and experiments. And that is a long and difficult way. Did you perhaps read some of the papers I recommended yesterday? There you will find some experimental evidence (as well as in the paper Marleen advised).

    But anyway, suppose the pufferfish skin pattern cannot be adequately explained by Turing models or, more in general by reaction-diffusion models, suppose … Than my question to you is: how would you explain this complicated pattern? Do you know any research about this question? What I want to say is that criticize a promising model without giving an alternative.

    ReplyDelete
  20. Leonardo vroeg "zijn de dieren, evenals de mens, eigenlijk ook wel zo geobsedeerd door patronen?"
    Ik weet niet of vrouwtjes zebra's geobsedeerd zijn door patronen van mannetjes, dat zou onderzocht kunnen worden, maar het schijnt dat vrouwtjes pauwen geobsedeerd zijn door de patronen op de pauwenstaart van mannetjes. De pufferfish: hebben mannetjes dezelfde patronen als vrouwtjes? Is er seksuele selectie? Dit kan in het aquarium onderzocht worden met proefmodellen die alternatieve patronen hebben. Of heeft het patroon een andere functie? Of is het een neveneffect van de werking van genen?

    ReplyDelete
  21. Marleen Gert Rolie,
    Natuur zal kunst nooit blijvend evenaren schreef de dichter (in het Italiaans). Die dichter was de kunstenaar Michelangelo die, toen hij dit opschreef, al zeker wist dat zijn “onnatuurlijke” kunst nog bekeken zou worden als hij allang tot stof was vergaan.
    Met het model van Turing weet de “kunstenmaker” – de hanteerder van het Turing-model – al bij zijn leven dat hij iets kan maken wat de natuur nog niet heeft laten zien.
    Ze zeggen dat modellen de werkelijkheid beter doen begrijpen. Een rare stelling: iemand die een gebeuren nog niet goed begrijpt maakt er een model van en zou dan kennelijk in staat zijn om dat model de natuur meer te doen evenaren dan zij/hij er van begrijpt.

    De klimatologen/ecologen maken modellen en ensembles van modellen, en daarnaast scenario’s van die ensembles van modellen die inmiddels iets over het jaar 2100 vertellen. Voor mij een reden om al dat modellengedoe een beetje sceptisch te bekijken. Het Turingmodel kan patronen tevoorschijn toveren die je in de natuur tegenkomt, maar ook patronen waarvan je maar moet afwachten of de evolutie ze sterk genoeg vindt om te surviven.
    Het punt is: de natuur weet niet van Turing patterns. De natuur doet maar wat. Dat wil zeggen: wat ze doet, lijkt nergens naar, behalve dat wij verwonderd zijn dat de natuur dat tevoorschijn tovert.

    Dat levert geen beter begrip op van de werkelijkheid. Hooguit ontzag, want waar een programmeur, of gebruiker van zo’n model, toch behoorlijk voor moet zwoegen, vooral omdat ie maar weinig tijd van leven heeft, lijkt de natuur zo uit de losse hand rond te strooien. En, ik heb het al eens eerder tegen Gert gezegd, de natuur heeft tijd zat.
    Als ik Gert goed begrijp is dat ook eigenlijk wat ie zeggen wil: voor ons begrip hebben we meer nodig dan de uitdraai van een modelberekening.

    In die modellen zit veel mathematica. En dat heeft er toe geleid dat geleerden hoe langer hoe meer (zijn) gaan denken dat die modellen de werkelijkheid zijn en dat de werkelijkheid slechts kenbaar is dankzij de mathematica.
    Ik denk dat de mathematica, dat wel calculus kent met differentialen en integralen, achter de werkelijkheid aanloopt. Archimedes zat in bad en zag zijn beroemde wet zich voor zijn ogen ontvouwen. Eureka. En als modellen gedraaid worden met behulp van statistische gegevens, wordt er gewoon gerekend volgens Meneer Van Dale Wacht Op Antwoord (in de goede volgorde dan).

    Ik moet nu aan een andere dichter denken: Joke van Leeuwen.
    Ze schreef het gedicht Van alles
    De eerste strofe begint met: er is nog van alles te doen.
    De tweede strofe begint met: er is nog van alles te zeggen,
    De derde strofe begint met: er is nog van alles te willen.

    En zo schildert ze in een handjevol regels het menselijke druk druk druk.

    Dan volgt de laatste strofe: de mol die spreekt.

    Niet te geloven, mompelen mollen,
    hoe volmaakt we gemaakt zijn voor
    achter- en voorwaarts bewegen
    en af en toe naar boven. Volmaakt
    ook onze ogen.


    Nou, en wie zou er nu een Turing pattern zoeken in zoiets onooglijks als een mol

    ReplyDelete
  22. Leonardo, bedankt voor je mooie proza en poëzie.
    Helpen modellen om de werkelijkheid beter te begrijpen? In mijn boek hanteer ik de kaar-landschap-analogie. Wat wetenschappers met hun modellen en theorieën maken is een soort ‘kaart’ om te helpen onze weg kunnen vinden in het ‘landschap’, d.i. de werkelijkheid waarin we leven: aarde, levensvormen, kosmos etc. Vaak zijn het goede en adequate kaarten, zonder die kaarten geen mobieltje of internet, maar ook geen waterleiding etc.
    Maar die modellen geven geen ultieme verklaring. Tenminste ik meen dat je zoiets bedoeld: wandelen of fietsen in een bos of langs de zee is iets heel anders dan de wetten van Newton of de theorie van Darwin begrijpen. Wetenschappelijke verklaringen kunnen interessante dingen zeggen over golfslag, eb en vloed, maar kunnen niet tippen aan de ervaring van de ondergaande zon aan zee.
    Het zoeken naar ultieme ‘verklaringen’ breng ik in verband met Iemand die deze wereld geschapen heeft en die zich bekend heeft gemaakt door Jezus Christus. Het zoeken naar adequate en ultieme verklaringen hoeft elkaar niet uit te sluiten. Misschien kunnen we het ook vergelijken met het verschil tussen luisteren naar goede muziek en het lezen van een partituur?

    ReplyDelete
  23. Marleen, your previous reference to a paper of Kondo led me to this review paper: 2021 - Studies of Turing pattern formation in zebrafish skin, by Shigeru Kondo.
    One citation: 'In this review, recent experimental studies using zebrafish are summarized for theoretical researchers who are
    new to biological pattern formation, and progress in theoretical studies based on experimental results is described.'

    Gert, maybe this is a nice introduction to pattern formation and Turing models?

    ReplyDelete
  24. nee Rolie,
    De aarde is niet geschapen. Het universum is er van alle eeuwigheid en zal tot in alle eeuwigheid bestaan. Ja, er wordt een figuur Jezus van Nazareth genoemd, op enkele plaatsen in de geschiedschriften, maar die kwam niet uit de hemel vallen en die is gewoon dood gegaan op aarde (terechtgesteld) en op een inmiddels onvindbare plaats begraven. Dus het verhaal kan niet gefalsifieerd worden, evenmin als het bestaan van een schepper.

    Ik heb de indruk dat je die boodschap kwijt wilde en ik zie ook geen antwoord in jouw reactie. Ik verwacht zo’n antwoord ook niet op een blog over evolutie, toch een voortvloeisel van de verlichting die we hebben mogen beleven.
    Dat gedicht stond daar ook niet voor niks. Wij mensen denken dat we een bizondere plaats innemen, in het universum zelfs: we zoeken wel naar leven maar niemand van die zoekenden is daar optimistisch over. Maar we zijn gewoon uit de evolutie voortgekomen, niet minder een wonder, maar ook niet meer een wonder dan die mol van Joke van Leeuwen.

    Dat was ook de strekking van mijn vraag aan Gert: wij zijn zo druk bezig met andere wezens en, in de woorden van Frans de Waal, we leren daar ook van. Wel, niet genoeg, want Frans de Waal meent dat hij de apen omhoog gebracht heeft (op de scheppingsladder?) en de mens een beetje zijn plaats heeft gewezen. Ik denk zomaar dat ie dat wel bij de apen “gevonden”, maar niet van de apen geleerd heeft.
    Dus wij mensen zijn erg tevreden met onszelf, en achten ons ook een evenbeeld van iets/iemand die het universum tot stand gebracht zou hebben, en dat mag je toch wel lichtelijk arrogant noemen. Zeker gezien de rotzooi die we aan het schoppen zijn in het Midden Oosten en Ukraine, en gelet op de panische angst die we hebben omdat het klimaat een beetje lijkt te bewegen, dat het eeuw in eeuw uit gedaan heeft.

    Ja die mol is niet ten onrechte tevreden met zichzelf. Hij heeft nog nooit een pufferfish gezien dus hij wordt ook niet afgeleid van zijn bestaan. En lijkt geen ambitie te hebben om daar wat aan te veranderen. Mol op de maan, zoiets. Misschien dat wormen en mollen elkaar een beetje in de gaten houden.
    Wij beschouwen ons als rentmeester op de eigendommen van een zelf verzonnen grootheid. En dringen binnen in de leefgemeenschap van apen om daar iets (on)zinnigs over te kunnen zeggen.

    Je maakt de vergelijking tussen muziek luisteren en een partituur lezen. Die zie ik niet zo. Als we muziek horen is de partituur er al. De partituur van de ketelmuziek die het universum ons laat zien en horen – denk aan big bangs, meteorietinslagen, zwaartekrachtimplosies in zwarte gaten – proberen we gaande de muziek te schrijven. En na zo’n 300 jaar verlichting zijn we nog niet veel verder gekomen en lijken we zelfs een beetje op de terugweg – nou ja, mijn observatie: back to the Dark Ages.

    ReplyDelete
  25. Thank you Rolie for the link to this interesting article.

    I'll just put it here again:

    Studies
    of Turing pattern formation in zebrafish skin


    It makes clear that the formation of stripes (at least in zebrafish) is better described by migration of pigmented cells than by reaction-diffusion mechanisms or the Turing model and this in 2021.

    That seems much easier for a biologist to follow. What surprises me is how nonchalant the authors are about the distinction between morphogens that are outside the cells, i.e. in the extracellular matrix, and morphogens that are inside the cells. The introduction mentions the egg cell that is said to be sensitive to morphogens. I have understood for a long time that the cytoplasm of the fertilized egg is not homogeneous and that this creates a distinction between the two daughter cells after cell division. This article states, in the introduction, that the fertilized egg, in contrast, is surrounded by a gradient of morphogens.

    The model assumes that a diffusible molecule (morphogen) is localized in the region of the fertilized egg. As the molecule diffuses, a concentration gradient is formed, and each cell in the embryo can tell its own location according to the concentration of the molecule.

    If this was the model until 2020, then they have been wrong about the egg for 70 years, unless they mean something in particular with “the region of the fertilized egg”.

    In my opinion, the inability to explain how reality works and the amazement that arises about it are not reasons to become or be a believer. You have certainly experienced for yourself how, as knowledge of the various physical and biological mechanisms increases, the admiration that exists for the world increases too. The better you can figure out how these natural scientific processes take place, the greater the admiration you have for the world we live in. That is why the sunset is extra beautiful when you know the natural properties of the sun and the properties of the atmosphere that make it turn red. I therefore disagree with how you separate the experience of a setting sun from science, from your own science. You are a scientist and as such you experience the world. John with the Cap experiences this very differently, perhaps less intensely, without wanting to be derogatory.
    The better we understand a person, the more we admire him and the more we love him. The more you know about a painting, the more you admire it and the more you love it.

    ReplyDelete
  26. Ik las in het document Studies of Turing pattern formation in zebrafish skin “Skin patterns are the first example of the existence of Turing patterns in living organisms.” Deze openingszin las alsof we zonder Turing nooit van patronen gehoord zouden hebben. Ik dacht, die patronen zijn al zo oud als de weg naar Rome, die kunnen toch niet door Turing ontdekt zijn? Turing zal toch niet de eerste zich verwonderende intellectueel geweest zijn?

    En ik ging op zoek, en vond een artikel over patronen bij huidziektes. Als patroon had het wel wat, maar niet zo mooi als de pufferfish, niet om naar te kijken.
    Hoe dan ook:
    - een Nederlandse anatoom, Bolk, die de head-zones ontdekte maar de eer gunde aan Henry Head.
    - een Oostenrijkse anatoom, Langer, die de Langer-lines beschreef, waar je zelfs een kerstboom in kon zien
    - een Duitse dermatologist, Blaschko, die de lines of Blaschko beschreven heeft: lijnen over de hele lengte van het lichaam, voor en achterzijde, in een zeer symetrisch verband.
    Allemaal begin vorige eeuw, of nog eerder.
    Wel, ik denk dat ook Griekse filosofen zich hiermee bezig hebben gehouden en hierover verwonderd waren.

    Maar goed, Turing ging op zoek naar: hoe kan dat?
    Meer mensen. En zo vond ik een verslag in THE CONVERSATION van Ankur Gupta, een assistent professor in biological engeneering over “How animals get their skin patterns”

    Daarin stond deze zin: Understanding how nature programs specific functions can help researchers design synthetic systems that perform similar tasks.
    Ik zet hem maar even bold. Want wat staat daar? Dat is toch Nature die functies programmeert. Dat is toch intelligent design?! En wij mensen kunnen leren dat te begrijpen en dan zelf die taak uitvoeren.
    Wat er niet bij staat is of we dat programmaatje dan ook in een eitje kunnen stoppen, of in een zaadje, zodat we als het ware kunstwerken op dierlijke huid kunnen produceren. Die het beestje dan door middel van voortplanting over de wereld verspreid.

    En ik moet denken aan dat verhaal van Roald Dahl, waarin een arme sloeber een schilderij in de huid van zijn rug laat schilderen door een beginnend kunstenaar die wereldberoemd zal worden, waardoor de rug van die arme sloeber goud waard wordt.
    Hoe dan ook, da’s nog eens wat anders dan Cees Dekker die een celletje probeert te bouwen.

    https://theconversation.com/how-animals-get-their-skin-patterns-is-a-matter-of-physics-new-research-clarifying-how-could-improve-medical-diagnostics-and-synthetic-materials-217035

    ReplyDelete
  27. Leonardo and Marleen, surprising how some thoughts I gave, give rise to so many others. Since I want to focus on the subject of this blog (pufferfish skin and plasma surface patterns), I keep my reaction short.

    Leonardo, I share your scepsis against synthetic biology. And further, I have deep worries about all that research trying to ‘make’ cells.

    Marleen , referring to the paper: “Studies of Turing pattern formation in zebrafish skin”, you write: ‘It makes clear that the formation of stripes (at least in zebrafish) is better described by migration of pigmented cells than by reaction-diffusion mechanisms or the Turing model and this in 2021.’
    Did you find this information in the paper of Shigeru Kondo et al.? There approach does not refer to migration of cells and does not deny the importance of the Turing model at all!
    On the contrary, Kondo’s model is a combination of so called kernels (an abstraction of cells) and the mechanisms of the Turing model. The latter summarized with: activation (positive feedback) on short distance and inhibition (negative feedback) on larger distance. Briefly called LALI: local activation, lateral inhibition.

    Marleen what you say about fertilized eggs, I know very little about that, so I can’t give a reaction.

    After the weekend I try to summarize what Shigeru Kondo describes in his review paper about zebrafish and pufferfish skin patterns. Surprising insights retrieved from the improved kernel-Turing-model and from experiments.

    ReplyDelete
    Replies
    1. One more remark.
      Marleen, about our common admiration about life and the cosmos. You wrote that the sunset would be extra beautiful to me because I know about scattering of light and nuclear fusion. For me that’s not the case. I forget about physics looking at the sunset. But looking at trees and flowers my admiration is indeed getting deeper when I realize the enormous complexity going in all the cells.
      And, please, when I say something about my believes, that is not meant to be something compelling you or anyone else to believe in a creator.

      Delete
  28. beste mensen: ik zat een lang weekend in een huisje in de natuur, en heb in die tijd niet mee kunnen doen met de discussie. Dus eerst al jullie comments lezen...

    ReplyDelete
  29. Rolie, a short comment.
    As far as I can see you did not react on this quote

    gert korthofWednesday, April 3, 2024 at 10:00:00 AM GMT+2

    Marleen thanks for the link to:
    Patterning, From Conifers to Consciousness: Turing’s Theory and Order From Fluctuations.
    https://www.frontiersin.org/articles/10.3389/fcell.2022.871950/full

    I found an important warning in the article:
    "For theorists, there have been disappointments along the way, in that patterns that appeared to match theoretical prediction were shown to arise by other mechanisms."
    (my bold)
    So, it is thus possible that a match between theoretical prediction and observed biological pattern does not prove that the theory is correct!
    This is an alarming and very important statement, and I think every scientist should take this very seriously. What is your reply?

    Furthermore, as far as I can see, you did not comment on the importance of falsification in science (the ability of scientific theories to be falsified, in contrast to verification, confirmation).

    ReplyDelete
  30. "For theorists, there have been disappointments along the way, in that patterns that appeared to match theoretical prediction were shown to arise by other mechanisms."

    Gert, referring to your note above of April 3th.
    Remarkably that you ask me to react on this one sentence:
    "For theorists, there have been disappointments along the way, in that patterns that appeared to match theoretical prediction were shown to arise by other mechanisms."
    You took it from a paragraph where the author (Thurston C. Lacalli) is summarizing what he knows about the application of Turing’s model, starting with: ‘Fifty years on (RB since 1970), Turing’s ideas have been successfully applied to a number of developmental systems … though the mechanistic details often differ from his original proposal, with chemical autocatalysis being replaced by other self-enhancing molecular or cellular processes, and distance effects by other means of material transport, or by mechanochemical effects (…).’ On the dots giving many references.

    So he emphasizes the great success of the Turing model, with modifications but always having the two basis properties of his model: self-enhancement (positive feeb=dbakc loops) and some kind of diffusion. Only then he talks about the disappointments, curiously without giving references.
    So why is your question so biased to disappointments? I would say, of course there are other mechanisms.

    Referring to the falsification, I have wrote about that on April 4.

    ReplyDelete
  31. Rolie, a further quote from Thurston C. Lacalli (2022) Patterning, From Conifers to Consciousness: Turing’s Theory and Order From Fluctuations:

    "Fifty years on, Turing’s ideas have been successfully applied to a number of developmental systems, though the mechanistic details often differ from his original proposal, with chemical autocatalysis being replaced by other self-enhancing molecular or cellular processes, and distance effects by other means of material transport, or by mechanochemical effects."

    Pay attention to "though the mechanistic details often differ" and further down: "This review is designed as a broad survey with a focus less on mechanistic details ..."

    So, Lacalli has a similar point of view as you have: it's Turing everywhere in any form with only some minor unimportant low-level mechanistic differences. That is the view of the theorist. However, the view of the evolutionist is concerned with the details: those mechanistic details probably have a genetic basis, and only genetic differences are inherited. So, from the point of view of evolution, those mechanistic details are all that matter, and Turing processes are taken for granted, they are assumed, they have no genetic basis, so can be ignored.

    ReplyDelete
  32. Gert, I read the main part of Lacallis paper and I doubt whether he is talking as a theorist. For example think of his 1981 paper: Dissipative Structures and Morphogenetic Pattern in Unicellular Algae.
    Your last statement is surprising to me: do you really think that ignoring possible morphologic processes may give a good account of evolution? In the evo-devo discourse researchers emphasize that you cannot understand evolution unless you understand developmental processes. So, you lose me again.
    On the other hand, I agree with you that all those biological processes leading to Turing patterns are not just mechanistic details.
    Later more about this question: How to find a more realistic biological Turing model?

    ReplyDelete
  33. Dear Rolie, of course evo-devo is important in evolution! What I mean is physical processes do not require biological heredity: there are no genes for diffusion and gravity, and so on. Evolution depends on heredity, and biologists can assume that physical and chemical laws operate as usual.

    ReplyDelete
  34. Rolie en Gert,

    Wanneer ik de studie van Kondo naast die van Rosenbauer leg
    Modeling of Wnt-mediated tissue patterning in vertebrate embryogenesis

    dan lijkt het erop dat er in beide gevallen met een model gewerkt wordt ( gebaseerd op Turing maar geavanceerder), dat uitgevonden is, nog voordat de moleculaire mechanismen ontdekt waren. Is dat de ‘verwondering’, te ontdekken dat het aan het menselijk brein ontsprongen model klopt met de realiteit? Ik begrijp namelijk niet goed wat er achter deze studies met modellen zit.
    De werkelijke patterning in dieren is zo complex dat je deze nooit echt zult kunnen ontrafelen. Je kunt er inderdaad hooguit een model van maken dat je van input voorziet om te zien wat er uitkomt en of je het programma goed hebt gemaakt door het met de uitkomst in de biologische wereld te vergelijken. En dan?

    In Kondo las ik het volgende in het abstract notabene:
    in real skin, it is not an alteration in concentrations of chemicals, but autonomous migration and proliferation of pigment cells that establish patterns, and cell–cell interactions are mediated via direct contact through cell protrusions. Therefore, the classical reaction–diffusion mechanism cannot be used as it is for modelling skin pattern formation.

    Dit lijkt me toch duidelijk, namelijk dat het Turing model niet werkt in dit geval.

    Dit artikel op Wikipedia (vers van de pers) over Wnt-signalering beschrijft in zo eenvoudig mogelijke termen de complexiteit van het effect van een morfogen, het morfogen Wnt. Het sluit mooi aan bij de observaties van Rosenbauer over de Zebravis. Hij ziet experimenteel, dat de morfogenen het patterning van het ruggenmerg en het encefalo induceren. Maar omdat de tijden waarop dit gebeurt te kort zijn, veronderstelt ook hij, net als Kondo, dat er cellen of filopodia van cellen (lange uitstulpingen ofwel de ‘protrusions’) zijn die met pigment en al migreren of dat ze pigmenten afleveren met behulp van de extreem lange filopodia. Beide auteurs zitten op een lijn.

    Wnt-signalering geeft eigenlijk een ‘aan’ en ‘uit’ signaal en lijkt niet onderhevig te zijn aan positieve of negatieve feedback. Het gen wordt afgeschreven of niet, al naargelang de aanwezigheid van Wnt op de receptoren van de cel. Dus het voorbeeld van Meinhardt & Gierer is eigenlijk te ingewikkeld. Zou de natuur toch eenvoudiger zijn dan de modellen en werkt zij met Occam’s Razor?


    ReplyDelete
  35. ‘Onmogelijke’ sterrenstelsels: hoe de James Webb-telescoop de kosmische jeugd van het heelal helpt inkleuren kopt de Volkskrant dit weekend.
    En als ondertitel ZO’N DUIZEND EXTREEM VERRE STERRENSTELSELS HEEFT RUIMTETELESCOOP JAMES WEBB NU ONTDEKT, FLARDEN VAN DE JEUGD VAN HET HEELAL. DIE STELSELS VERBAZEN ASTRONOMEN NOGAL EENS. WAS DE JONGE KOSMOS SOMS HEEL ANDERS?
    Het artikel opent dan met: Ze kan zich er nog dagelijks over verbazen, zegt sterrenkundige Pratika Dayal van de Rijksuniversiteit Groningen. ‘We kijken nu dieper het heelal in dan ooit eerder in de geschiedenis van de mensheid. En wat we daar zien, de kwaliteit en hoeveelheid van onze meetgegevens, overtreft nu al onze stoutste verwachtingen.’

    Het heelal wordt uitsluitend “begrepen” vanuit modellen … totdat we weer een stukje dieper kunnen kijken.

    Inmiddels is er een wetenschap die zich baseert op ensembles van modellen met daarop gebaseerde scenario’s en ensembles van scenario’s die probeert iets te zeggen hoe de wereld er over 100 jaar uit zal zien. Daar komt de klimatologie aan te pas, met ogenschijnlijke diepe inzichten in het weer, maar van de wolken en van de zon begrijpen we nog weinig tot niets. En daar komen alfa wetenschappen aan te pas, van economie tot sociologie,

    Ik denk dat die wereld van modellen en modelleurs op een groot misverstand berust: dat modellen je de werkelijkheid zouden doen begrijpen. Terwijl modellen de uitdrukking zijn van hoe de modelleur vandaag denkt de werkelijkheid te kunnen interpreteren. En zo heeft Turing niet veel meer gedaan dan gezegd dat het niet onmogelijk was dat de natuur mooie vormen en structuren kon produceren – “want ik kan het ook met een computer”. En hij kon het: verander één parametertje en je ziet weer iets heel anders … op het beeldscherm, wel te verstaan.

    En dat dringt bij mij een vraag op: is het denkbaar dat symmetrie, “mooivormigheid” – dat toch ook iets zegt over eenvoud - ook sturend is bij natuurlijke selectie en survival? Dat lelijkheid, misvormigheid, niet in staat is om zichzelf in stand te houden?

    ReplyDelete
  36. Marleen, bedankt voor het interessante artikel van Rosenbauer. Het artikel laat zien dat vrije diffusie zowel voor het Turing- model als voor het Wolpert-Franse-vlag-model uiterst belangrijk zijn. En dat deze vorm van diffusie wel eens te traag zou kunnen zijn om de embryonale groei van hersenstructuren van zebravisjes te verklaren.
    Maar als transport van morfogenen verloopt via lange uitstulpingen (cytomemen) van het celmembraan dan kunnen morfogenen zich sneller en ook gerichter verspreiden naar andere cellen.
    Dat voorstel van Rosenbauer sluit inderdaad aan bij het model van Kondo. Maar er is een groot verschil. Rosenbauer gebruikt een Wolpert-model uitgebreid met cytomemen, Kondo een Turing-model met cytomemen.
    Dus dat je bij Rosenbauer geen feedback loops vindt is logisch, want die zijn kenmerkend voor het Turing-model, niet voor het Franse-vlag-model van Wolpert.

    ReplyDelete
    Replies
    1. Hierboven: "vindt is logisch" moet natuurlijk zijn "vind ik logisch"

      Delete
  37. Gert, of course physical and chemical laws don’t require genetic memory. But that’s not what you said before (April 13). There you said: ‘… from the point of evolution … Turing-processes are taken for granted, they are assumed, they have no genetic basis, so can be ignored.’
    We are talking about evolution theory and to find a proper theory you cannot ignore any chemical or physical law of process, including Turing-processes. During this discussion about circular causality I proposed that organism and their genes only can operate within that lawful frame, including Turing-processes. And therefore genes get the basics for patterning processes for free – genes don’t invite but only use what is in general available. So, indeed evolution depends on heredity AND on all what physical and chemical laws may give.

    ReplyDelete
  38. Leonardo schreef: "is het denkbaar dat symmetrie, “mooivormigheid” – dat toch ook iets zegt over eenvoud - ook sturend is bij natuurlijke selectie en survival? Dat lelijkheid, misvormigheid, niet in staat is om zichzelf in stand te houden?"
    Symmetrie is zeker belangrijk in evolutie, maar niet vanwege Plato! Een dier met drie poten is onhandig, alle dieren (klein en groot) met poten hebben evenveel poten links als rechts. Dus symmetrie om down-to-earth laag-bij-de-grondse redenen. Het werkt gewoon beter. Ik denk dat ons schoonheidsbegrip uiteindelijk daar op gebaseerd is...

    ReplyDelete

Comments to posts >30 days old are being moderated.
Safari causes problems, please use Firefox or Chrome for adding comments.