Mathematicians Visualize Evolution in Algorithms

Molecules just want to come together
and become living cooperators,
say these university eggheads

 

— The window they claim to have opened is more like a crystal ball —

What is more random than the Stuff Happens Law? What is more unguided than molecules bouncing around in a primordial soup? Aha, but what is more deductively certain than mathematics? Here, Darwinians try to latch onto the prestige of math* to lend support for evolution and the origin of life. Will it work?

*We use the American form “math” instead of the UK form “maths” that is awkward to pronounce and confuses singular and plural forms, making sentences sound grammatically incorrect, such as in “maths unlocks molecular interactions.” We hope the Brits will find the American usage more sensible.

Maths unlocks molecular interactions that open window to how life evolved (University of Queensland, 27 April 2023). The smiling lady photo at the top of this press release lends an immediate mood of credibility to what seems to be an audacious claim below it.

Landmark research published in Nature Communications by mathematicians Dr Robyn Araujo at QUT and Professor Lance Liotta of George Mason University in the US sets out the definitive picture of biological adaptation at the level of intermolecular interactions….

“Until now, no one had a general way to explain how this vital process was orchestrated at the molecular level through the vast, complex, often highly intricate networks of chemical reactions among different types of molecules, mostly proteins.

“We have now solved this problem, having discovered fundamental molecular-level design principles that organise all forms of biological complexity into robustness-promoting, and ultimately, survival-promoting, chemical reaction structures.”

But What Do They Mean?

Understand that their view is far from intelligent design. The “vital process was orchestrated” not by a mind or intelligence, but by material laws of nature following mathematical forms. The “design principles” were not designed; they just are. In short, molecules come together naturally and live. They want to. They have to. Indeed, they are “engineered” to do so by some unspecified, blind engineer:

“These complex intermolecular interactions must implement a special type of regulation known as integral control – a design strategy known to engineers for almost a century.

“However, signalling networks in nature are vastly different, having evolved to rely on the physical interactions between discrete molecules.  So, nature’s ‘solutions’ operate through remarkable and highly intricate collections of interactions, without engineering’s specially designed, integral-computing components, and often without feedback loops.

Engineering without an engineer? What a thought. How can that work in solutions of mindless molecules bouncing around? The press release uses a bit of personification to make the notion palatable:

“We show that molecular network structures use a form of integral control in which multiple independent integrals, each with a very special and simple structure, can collaborate to confer the capacity for adaptation on specific molecules.

Now there is collaboration without collaborators, and conferring without a conference of minds using intelligence and wisdom to make decisions on what should be done.

Solving a Grand Challenge

Liotta is proud of his part in “the quest to uncover the fundamental design principles of biological systems throughout nature” which he considers “to be one of the most important and far-reaching grand challenges in the life sciences.” He and Araujo have boosted their universities’ prestige with their new algorithm they call RPA: Robust Perfect Adaptation. Wow. An acronym!

“On the basis of this ground-breaking new research, RPA currently stands alone as a keystone biological response for which there now exists a universal explanatory framework.

“It’s a framework that imposes strict and inviolable design criteria on arbitrarily large and complex networks, and one that now accounts for the subtleties of intricate intermolecular interactions at the network microscale.

As is customary, the press release includes a statement that this achievement might also help cure cancer.

“At a practical level, this discovery could provide a completely fresh approach to tackle grand challenges in personalized medicine such as cancer drug resistance, addiction, and autoimmune diseases.”

So there’s lots of pomp and ceremony in this press release. With the hype over, what does the actual paper say?

The Hard Parts

Universal structures for adaptation in biochemical reaction networks. Araujo and Liotta, Nature Communications 14, Article number: 2251 (20 April 2023).

The authors’ audacity continues from the press release into the paper.

Although RPA-conferring CRN structures are known to be quite subtle, the framework we present here delineates the fundamental principles fully. Only a complete and truly general picture of the integral control problem in CRNs, as we present here, can demarcate the evolutionary trajectories along which complex adaptation-capable biological networks can arise from simpler building blocks, and provide a roadmap for either preserving or disrupting the RPA property in natural, diseased or synthetic networks through design alterations or pharmacological interventions.

Wait a minute. How can they come up with a “universal” adaptation algorithm that presumably affects the origin and evolution of life by studying only existing living systems and synthetic (intelligently designed) systems? Isn’t that begging the question?

Here we identify the universal principles by which all possible instances of RPA-capable CRNs – in all living systems on Earth, as well as in synthetic biology – construct internal models of any possible stimulus or disturbance, or change in total molecular abundances, thereby allowing the CRN to implement integral control.

Theory La-La Land vs Real World Biochemistry

There’s no serious discussion of how randomly mixed chemicals in a primordial soup could “adapt” in any purposeful way and become alive. There’s no discussion of entropy, the Second Law of Thermodynamics, or the many problems that are routinely ignored in origin-of-life (OOL) scenarios such as investigator interference, implausible starting conditions, linkage of unrelated scenarios, achieving sequence specificity, avoiding damaging cross-reactions, or (especially) overcoming immense probability bounds (see online book and Illustra video).

In fact, despite the press release claim that the work will “open [a] window to how life evolved,” this paper glosses over life entirely. The authors live in theory-land without offering illustrations of what really goes on to make molecular machines and functional information. Read the paper yourself and see if this is not so. The few cases where protein interactions are mentioned are either simplified cases or theoretical ones if unspecified regulators are operative.

The bulk of the theoretical work is in the Supplemental Information document, which never mentions real world molecules like DNA, RNA, lipids, sugars or proteins, nor does it mention entropy or any of the other problems with OOL studies. It appears to be an abstruse model of handwaving in theory-land with no real world application to the origin of life or evolution.

So did they demonstrate evolution by mathematical algorithms? Consider this paragraph, and remember that they want to find chemical reaction networks (CRNs) that exhibit RPA (Robust Perfect Adaptation) like those at work in living cells.

Here we briefly review again the analytical approaches that can be used to check whether a CRN at hand has any potential for RPA capacity through an analysis of deficiency (and a prior decomposition into independent subnetworks, wherever possible). Nevertheless, it’s worth always bearing in mind that one should have a reason to suspect that a CRN exhibits RPA. It is clear from our analysis of RPA-capable CRNs at the network microscale, and from our topological analysis of RPA-capable network architectures at the network macroscale, that RPA-capable networks are actually very ‘special’, with very specific structural requirements, and are therefore extremely rare in the space of all possible networks. If one has a completely arbitrary network, with no particular reason to suspect it might be able to exhibit this special type of robustness, then it almost certainly doesn’t.

The upshot is this: even if living systems and synthetic systems exhibit RPA, that has nothing to do with how they originated.

No further questions, your honor.

Verdict: Guilty. This paper is about intelligent design, not evolution.

Recommended Resource: Watch the newest “Long Story Short” video “Reality Check” about misleading origin-of-life claims.