Thanks Phil for the comments. I just now looked to see if I had any responses, and I was pleasantly surprised. Thanks to all who offered comment. I’m tired and a little under the weather tonight, so I’ll have to ruminate and rest on the responses before I respond more fully. My first reaction is that those who challenged me seem to not appreciate the concept of coordinated change. But maybe I’m missing something. I look forward to engaging and learning from you guys.
Ray
It’s a simple concept and we understand it perfectly well. It happens all the time in biology, and it’s not an issue for evolution. There are many reasons why, and robustness is just one of them.
You’re missing numerous things. But it doesn’t have to stay that way.
Great. Looks like I’m going to learn something.
Let’s start by getting some clarification to ensure we’re talking about the same thing.
In the SE world that I know we employ “Change Management” to ensure that the many interconnected sub-systems that are part of the larger “system-of-systems” retain their ability to interoperate both intra-system and inter-system. This is no small task. When proposing a new change to a “system-of-systems” we first identify the functional change and disseminate it to all the other sub-systems that have the potential of being impacted. The functional changes are reviewed by the systems manager for each of the sub-systems to determine if there will be an impact. Clearly, if there is an impact, the “system-of-systems” will fail unless all the impacted sub-systems are not modified to accommodate that change. The sub-system changes necessary may be minor or major requiring a complete redesign of the sub-system across all dimensions of its functionality. The changes required typically fall under one of four categories: 1) Command, 2) Control, 3) Communication, 4) Computation (processing) (C4). Each of these 4 categories require common timing and spatial location functionality to ensure they all can operate in synchrony. As I said above, this is no small task. We can hope that if we change one small part of one sub-system, the other sub-systems as well as their C4 functions will all change on their own once the change goes into effect (maybe this is what you are referring to as “robustness.”) The other approach is to engineer the changes required across all the sub-systems and their C4 functions. This, of course, requires careful design. Note that design always begins with a vision of the final block of functionality intended for the “system-of-systems” with it new change(s). From that vision, designers work their way down from the highest layers of function to the lowest to arrive at a complete set of system specs. Those specs are then compared to the specs of the unchanged original “system-of-systems” to identify which sub-systems must be changed. But wait, since each sub-system may change, its change may impact the ability of other sub-systems to continue to interoperate so that the “system-of-systems” function(s) will be realized. Consequently, there are typically many rounds of coordination among the sub-system managers to arrive at a set of stable change specs. This is an involved process, that requires industrial strength abstract thinking. Do cells have the ability to do abstract thinking?
This is my understanding of the SE process. It’s universally accepted and practiced across nearly all professional disciplines, especially the US Government. It relies heavily upon physical laws and constants. No exceptions and no magic allowed.
When I look at the operation of the cell which is clearly a “system-of-systems” which itself is part of a larger “system-of-systems” (organs and organisms) with serious C4 function, I’m trying to understand how it could have evolved. Evolution begins with a random change of one or more nucleotide bases that, in turn, result in the production of one or more proteins that fold in one or more ways to enable it to perform its function. The problem from an SE perspective is that if there are 1000 other proteins in the cell that must also change in specific ways (not random), how does that one change propagate forward to bring about a beneficial change? How do the other 999 ensure they change exactly as required to ensure the first random change results in a viable cell?
I appreciate your comment that responded to my original post, viz. “It’s a simple concept and we understand it perfectly well. It happens all the time in biology, and it’s not an issue for evolution. There are many reasons why, and robustness is just one of them.” Let me respond with:
Looks like I still need some help. Thanks for being patient.
Ray
You made no mention of robustness or flexibility or redundancy, all of which I know to be highly relevant in SE. I have some slight literacy in this area, but one need hardly be an engineer or a control theorist to know that resiliency and robustness are critical matters. I’m not sure you have represented your example accurately.
Which might not matter, because…
This is what we call a strawman. You have created this strawman inadvertently, I suspect, but it is a strawman nonetheless. In all of your comments so far, you have started with a picture of protein-based systems in a cell that is inaccurate. Your picture omits flexibility, redundancy, and robustness. Proteins are famous for all of those things. Your picture omits known mechanisms of compensation, some of which are built into individual proteins but most of which are properties of the network. Your picture is almost certainly a caricature of human engineering (or so I must hope, because the brittleness you envision seems crazy bad to me), but what matters is that your picture is so far removed from what we know about proteins and cells that it’s not really worth any further discussion.
Maybe it was designed in. So?
I’m not saying that. I think you are really stuck, as in almost completely hamstrung, by your inaccurate picture of what proteins and cells are and do. Your question is nonsense because you are already assuming that hundreds of proteins have to “change in very specific ways” when one protein is changed. We have already told you that a simple reflection on existing protein diversity, between you and your relatives, would reveal this to be false. If you can’t make that connection, then there’s not much more to discuss with you.
Stephen,
Whew! Looks I got you spun up. I didn’t mean to make you angry. What I’ve told you about SE is rock solid, proven and universally accepted. It’s clear that you don’t understand SE which is apparent from the things you’ve said. As in most debates, folks tend to talk in terms that the other side doesn’t understand. I ask your patience so we can make some progress to find out where we really can come together. The SE principals are true as I’ve said. I clearly don’t know biology very well, but I know that biological systems don’t violate physical laws and principals. My quest is to understand biology well enough and within the constraints of those laws and principals so that I see the merit in the evolution argument. Currently, I see evolution violating those laws, mostly from looking at the stochastic models. Actually, when considering SE principals, I think the models underestimate the improbability of evolution. I know the laws and principals of physics and mathematics, but I don’t know how biological systems use them. You use terms like robustness, protein diversity, flexibility and redundancy. These are well understood terms/phrases in SE and can be explained easily in an engineering context. Help me to understand what you mean by them in the context of cell biology. They should be the same unless magic is going on. I’m sure we can find that explanation, if indeed evolution is true. I’m willing to work with you to understand biological systems and see how they are consistent with SE principals. As I said, if evolution is true, SE will provide a solid way to verify it. If evolution is not true, SE will expose it as fraudulent. If evolution is not true, your use of terms like robustness, diversity, flexibility, and redundancy will be exposed as merely terms used to hide the fact that you really don’t understand what’s going on in the cell and how it accomplishes its extraordinary engineering feats that so far are orders of magnitude greater than anything mankind has designed. But if you can bring me to understand how the cell does it, I’ll be forever in your debt.
A good place to start might be your statement “Your question is nonsense because you are already assuming that hundreds of proteins have to “change in very specific ways when one protein is changed. We have already told you that a simple reflection on existing protein diversity, between you and your relatives, would reveal this to be false.” This statement needs from justification. You appear to be claiming that a protein can change randomly and that other proteins that depend on its function will continue to function the same way or in improved ways. That’s patent nonsense. If proteins are interdependent in their function in the cell, and one changes, it will impact the function of the others that depend on it by definition. Are you saying that the proteins in the cell are not interdependent? Are you saying that there are other proteins laying around in wait so that when one protein starts to function in a new way, the dormant ones will pick up the slack or even make all the other proteins work more efficiently or differently in a beneficial way? I would love to see the Command and Control function that makes that happen. Stephen, that doesn’t even pass the giggle test! Clearly I’m missing something. Help me understand.
Ray
What physical laws are you thinking about here? I’m a physicist and can affirm that evolution definitely does not violate any laws of physics but perhaps there are some others you have in mind.
Going back to your first post, do you have evidence of this claim? Examples or experiments that others have done to demonstrate that it is impossible to change any proteins at all (with the assumption they must have been created in their present form?).
I’ll post some articles to get you started. But it’s a little worrying that you didn’t try to find them yourself. If you are “on a quest to understand biology well enough” then you should already be able to explore the basics of evolutionary theory and the basics of protein structure and function. Some of that information is right here in this thread (earlier, in the discussions of IDPs). So, below are some key articles that should help you understand some protein evolution-related concepts, but it’s up to you to do the work to figure them out. And in my opinion, it is inappropriate and a waste of my time to exchange posts with you if you do not intend to learn more.
Robustness in biology/evolution:
Andreas Wagner is one of the world’s best and his book The Arrival of the Fittest would be enlightening and enjoyable to anyone truly interested in evolution from a systems point of view, and especially to anyone interested in various aspects of biological resiliency. An article-length review is The role of robustness in phenotypic adaptation and innovation. Specifically on protein structure, and in fact on the kind of protein structure that seems at first glance to be stringently controlled and less tolerant of change, see The Robustness and Evolvability of Transcription Factor Binding Sites
A great review of robustness and evolvability by someone I know pretty well, also a world expert:
https://www.cell.com/trends/genetics/fulltext/S0168-9525(10)00110-1
Here is very recent review, focused on the level of development (neural development, my favorite). I can provide the PDF on request.
https://www.cell.com/trends/neurosciences/fulltext/S0166-2236(18)30154-1
Variation in proteins (“protein diversity”)
This one should require no sources, since IMO everyone should know that diversity in individual organisms implies diversity in genes and therefore (at least part of the time) in proteins. But it’s worth noting that despite the fact that protein variation has been studied since the early days of molecular biology (here is a historical review of some classic papers from the 1960s), we have only recently (last decade or so) been able to catalog variation in genomes and genes and proteins with high throughput. Specifically in humans, the Exome Aggregation Consortium is cataloging protein-coding variation in large cohorts. Their first edition was published a couple of years ago.
The short summary is that variation in protein structure is common within populations, and this should be obvious on reflection to anyone who knows the tiniest bit about what proteins and genes do. If protein structural variation were not tolerated, then you would not be reading this.
Protein flexibility
I refer you to the preceding discussion of IDPs, but also to our knowledge of chaperones, “moonlighting,” functional switches induced by post-translational modifications, to name a few big ones that came to mind. Proteins are flexible in how they respond to changes (mutations), which is a form of robustness, and they are also flexible in what they can do. This, of course, is a generalization; tolerance of change varies over orders of magnitude. The point is that if you propose that a change in a protein must necessarily cause problems or precipitate network-wide adjustment, then you don’t know what you’re talking about.
Redundancy
Redundancy at the protein level was actually a somewhat vexing problem for biologists (like me) studying knockout mice in the 90s and 00s. (It’s still a problem but less vexing.) We would make a knockout for the purpose of studying the roles of a protein in development or physiology. The protein was known to do interesting things, so it was a candidate for the long and arduous process of making a knockout. And then at the end of the line… there would be no phenotype. No effect. Why? Big question, but in many cases the answer was redundancy. Protein X did in fact control process Y, but another protein could stand in when necessary. Maybe the backup is a closely related protein, and maybe it’s not so closely related. In some cases we still don’t even know how the deficit is made up. But the take-home message was clear: in many cases, there is functional redundancy in the system, such that removing an entire component won’t necessarily bring the system down.
In plenty of other cases, we know in advance that there is redundancy, functionally and/or structurally, in a biological system. It’s just one more reason why no biologist would suspect that a change to a single protein would necessarily be a big deal.
It’s actually fact.
Let’s stop here, because I think it is doubtful that we will have a discussion. In multiple places in your post, you dismiss biological knowledge as nonsense, you suggest that biologists (me in particular) are using fancy words to hide ignorance, you implicitly suggest that evolutionary biologists are dolts who can’t see obvious weaknesses in the theory. You are mocking the science, while showing us that you have no understanding of what it is. You ended with “help me understand,” but I doubt you want to understand. So. Please don’t expect me to walk you through the basics of protein function, or evolutionary biology. That’s your responsibility.
Thanks Matthew for asking your questions. My thoughts in saying that I see evolution violating those laws stem mostly as I said from looking at the stochastic models. Clearly what’s going on inside the cell doesn’t violate the known physical laws. What I was saying is that the way I see evolution and the reality of what’s going on in the cell just from an SE perspective, evolution can’t happen. It’s too unlikely. You’ve seen the Model that Stephen Meyer uses where he takes Doug Axe’s number of 10&64 along with the chirality (10&45) and peptide bonding (10^45) problems to show that the odds of getting a functional protein that can support life to be 1 in 10^154 for a single protein of 150 amino acids in length. However, when you look at the complex structure of the cell and its large set of interdependent proteins working together from an SE perspective that recognizes a layer of complexity that Meyer, et al. haven’t incorporated into their model, evolution (random change and natural selection) becomes even less likely. One area that gives SE’s like myself challenge is that of communications. Entities that communicate must worry about signal to noise ratios, data formats, data speeds (bandwidth), error detection and control, modulation compatibility, etc. Getting all these right by design is hard. Getting them right by chance/random mutations takes my breath away. One aspect of the communication function going on in a cell that makes me scratch my head is bandwidth. When you consider the many functions going on in the cell that depend on ingesting and processing data to decide what to do next, when to do it, where to do it, what to do when done, it suggests a lot of data is moving around the cell and at very high speeds. Also, a lot of processing must be going on. Where’s that processing function happening? Where are results both intermediate and final being stored even if only temporarily? Is there a central clock that allows processes to kick off at specific times or are all processes depend on signaling from its environment (other proteins) and there’s a well establish order so that when one process finishes, it triggers the next process by perhaps a different protein or group of proteins? Is it like a setup of dominos? You push the first one to start, and they all fall in order. The cell is clearly not a Von Neumann machine. It’s probably one that has a distributed architecture where computation, storage, and communication are distributed which is much more advanced. Has anyone done research in this area?
I’ve been thinking about the navigation problem inside the cell. That alone requires considerable bandwidth and data processing. Our GPS system in our cell phones is a good analog. The “system” must know where you are and where you want to go. Then it calculates the best path and options giving you the times estimated given other conditions such as traffic. I would think that in the cell, not only does it generate the basic navigation information for the protein doing the traveling, but it sends it to the terminating site so that it knows something is coming and when it will arrive. What triggers it to start its trek? How does it know it has arrived? How does it know what to do when it arrives? What kind of signaling is used in each step? It’s hard to imagine that as being all chemical? What’s the addressing scheme for all sites in the cell? What metrics are generated and how are they fed back to the Executive Controller in the cell so it can makes changes to guarantee the health and proper function of the cell? By the way, how is executive or central control achieved in the cell? Look at an Amazon warehouse and all the internal functions going on. Communication among all the components (men and machines) is high volume and requires a strict and disciplined execution. The cell dwarfs an Amazon warehouse in its complexity, and yet we are to believe that it evolved. If that’s true, I want to believe it. But right now, that’s a challenge. If it was really designed, I want to believe that. I don’t want to believe something that’s not true. I’m willing to consider both sides to help me get it correct.
I also don’t want to offend anyone. Stephen seems annoyed with my questions, but he also seems resistant to challenge no matter how compelling the challenge may be. He appears to not have experience or an in-depth knowledge of SE. Those who have never forget it, and the experience leaves scars. But for sure, SE is solid in its science and it doesn’t give partial credit on exams. Right now, SE is telling me evolution is a non-starter. Nonetheless, I want to understand your thinking. What are you seeing that I’m missing?
Regarding your second question about my original post, my point is that at the “system-of-systems” level where all the sub-systems are interdependent, changing one without respective changes in all the other sub-systems will be destructive. I may have overstated how sensitive the “system-of-systems” is to changes in a single sub-system. We often will do failure mode analysis to determine the threshold when the full “system-of-systems” as a single or group of sub-systems start to fail, and often the threshold is not zero, but sensitive systems will fail quickly when an important sub-system begins failing. The point to take away is that coordinated change is the way “system-of-systems” evolve, and it’s by design. To get a new species, there are too many significant changes that must be made in coordination at the same time. Laws of probability show that to be extremely unlikely. Random mutation and natural selection won’t do it for a complex “system-of-systems.”
You’re right. I was way to strict in the way I used those terms. I apologize. I clearly need to read more and get a better understanding of biology. Thanks for those references.
The essence of the problem I’ve tried to articulate is simply that when you have high degrees of dependency among sub-systems (proteins), and one changes without commensurate changes in those other sub-systems that depend on it, the larger system-of-systems that depends on all of them will begin to fail. Expecting commensurate changes to all the other sub-systems randomly gives me serious pause. How do you rationalize that? Even if designed, it’s hard. Expecting all the other 999 sub-systems to randomly acquire the needed changes at random and all at the same time as the one that mutated is really hard to accept. but I’m open to hearing how you can believe it.
Thanks again for your patience. I’ll try to do better.
Ray
Hi Raymond,
I’d just like to pick up on something that @sfmatheson said here:
Reading your post, it seems to me that your understanding of software engineering, while strong, is somewhat traditional – certainly when you talk about subjects such as Change Management, it suggests to me that you’re used to rolling out large upgrades on a relatively infrequent basis (weeks to months), following very cumbersome and bureaucratic processes to do so. What I’m wondering here is, how familiar are you with the latest innovations in software engineering and application lifecycle management – concepts such as Continuous Delivery, microservices, and canary builds? I’d never heard of INCOSE before, but reading some of the literature on the link that you provided, it all seems very abstract, theoretical and academic, without much reference to real-world experience, to the extent that it’s difficult to even figure out what they’re actually talking about. In some parts of the industry, people such as that are referred to as “architecture astronauts.”
I don’t know much about protein folding and all that, but I can tell you that a lot of modern software development techniques are actually inspired by biological evolution. Take, for example, the concept of “canary builds.” These days, modern tools and platforms such as Docker and Kubernetes allow you to put each individual component in your “system of systems” onto a collection of several load-balanced containers. When applying a change (think of it, if you will, as a mutation) you first roll it out to one or two “canary containers” which are then heavily monitored. If the change is successful, it’s then rolled out to increasingly more containers, whereas if it isn’t, it’s just quietly rolled back and discarded (think of this, if you will, as the changes in allele frequencies that the “official” definition of evolution refers to). Another, related tendency is to move away from large, occasional, high-risk changes and move towards small, frequent, low-risk changes instead – the idea being, “if it hurts, do it more often.”
A good book to read on the subject is “Building Evolutionary Architectures” by Neal Ford, Rebecca Parsons and Patrick Kua:
Of course, the changes (mutations) involved in software development aren’t random by any stretch of the imagination – you wouldn’t just develop software by randomly flipping bits in your Git repository – but the parallels are still very much there. I’d be interested to know if these particular concepts have any bearing on your thoughts on the matter.
Consider a simple case. Suppose you inherit a genetic mutation that causes you to be taller because your leg bones grow longer. This was originally a mutation, a change to one subsystem without any commensurate changes to the other subsystems that depend on the bones and on leg length (muscles, nerves, blood vessels). Do you think your body will begin to fail? Do healthy people of different heights exist?
I think you’ve made that clear enough. And you’ve been told multiple times that your scenario is not an accurate picture of reality with multiple examples just given by two real live biologists (@sfmatheson and @glipsnort) that demonstrate why your scenario does not represent reality. If your scenario was true, then yes, we should all have serious pause that the mechanisms we presently understand are inadequate to explain how proteins can ever change. Even if it that was the case, not being able to fully explain something doesn’t mean its evidence of a supernatural being. It just means there is more science to be done! But your scenario is not true so to continue arguing such is feeling a bit like a whirlwind.
When I told you that was a strawman, and repeatedly pointed you to obvious refutations of your premise, I was expecting you to infer that neither I nor anyone else believes that. Good luck with your research. You will not learn anything if you approach the topic with a strawman as your starting point.
Don’t take it personally. And you don’t need to apologize for anything either. Some people get a little touchy because of their past frustrating experiences with ID-proponents.
Do you really think a single mutation will cause legs to grow longer? Or maybe its is a whole array of changes designed carefully and implemented partially in response to the environment? I actually want to understand you guys and your thinking. If I’m wrong, I’m open to being corrected. You definitely got me on my being overly strict in my example. I have some reading to do per Stephen, et al. Thanks for your patience. I have some work to do, but I’ll be back.
Thanks Steve. I’ll read it soon. By the way, has anyone in this group reviewed the new ID Book, “Theistic Evolution?” Remember, I’m reading both sides. The most interesting part of this effort is to see the response by the other sides. It’s often very revealing. Also, I’m taking your criticism about my SE ideas as a strawman seriously, and am trying to see how it applies. Not seeing it yet, but still thinking about it. Speaking of logical fallacies, how is your quote below not “begging the question.” More later…
Do you mean the book edited by Moreland? It was discussed in this thread
My take is it sets up a strawman version of EC and then knocks it down. There have been several threads with discussions with Ann Gauger and Richard Buggs among others. So we do get the ID folks giving us their sides here.
I’ve been pretty specific about the strawman, and so have a few others, so you might be the only one who can’t see how it applies. But let me try one more time. The strawman that is hindering your ability to understand evolution is your apparent belief that a change (mutation) in a single protein will create a systemwide disruption that requires simultaneous changes in hundreds or thousands of other proteins. You have maintained this strawman through at least 3 rounds of response. Here is where you most recently erected it:
when you have high degrees of dependency among sub-systems (proteins), and one changes without commensurate changes in those other sub-systems that depend on it, the larger system-of-systems that depends on all of them will begin to fail. Expecting commensurate changes to all the other sub-systems randomly gives me serious pause.
Your flawed premise is that changes within subsystems cannot be accommodated by the larger network. You then make your error quite a bit worse by claiming that evolutionary biologists actually believe that genetic change over time involves “commensurate changes to all the other sub-systems randomly.”
Your premises are false. They require no further discussion because they’re ridiculous. Worse, in this case, it seems you are projecting false beliefs onto me and others. Because there is no other way you could believe this:
First, it’s your job to explain why this is “begging the question,” but you needn’t bother. For me to have begged the question, I would have had to assume something that is incorporated into the question or problem I was addressing. What I did was straightforward and has been echoed by others here. I simply took your premise (that a single change in a protein will cause widespread failure) and applied it to what we actually know. I chose the simplest case I could, but we know a whole lot more detail about mutations and genetic diversity. The case was this: if your premise is true, then protein-coding mutations should make it almost impossible to reproduce. Why? Because they happen, regularly. They’re not some theoretical idea floating in the brains of people without engineering degrees. We measure them on vast scales. No one who understands the slightest bit about genetic diversity, including protein diversity, would ever begin with the premise you begin with.
So, let’s recap. You have now induced me to reply multiple times to a silly strawman that was barely worthy of a single response. And you have labeled a simple refutation by me a “fallacy”, inducing me to explain why it obviously is not. It’s obvious I need help (at least on weekends) budgeting my leisure time, but perhaps you will understand my claim now that you are obligated to show evidence that you are trying to understand the very VERY basic principles that we are discussing. I’m much more interested in talking about control systems and design and their evolution, which is an extremely interesting topic (Andreas Wagner is one important voice in that research field) than I am in explaining rudimentary principles of basic biology to people who start their comments with condescending giggles about ideas they don’t remotely understand.
Yes, absolutely – that kind of phenotypic effect is completely standard in biology.
No, it can’t be a whole array of correlated changes. This kind of mutation is part of the within-species genetic variation that we’ve been talking about. If only a whole array of correlated changes could be successful, then the whole thing would fall apart after the first generation, since different genetic variants are inherited independently.
This is one example of a kind of robustness to variation. Robustness is necessary not only for evolution, but simply for the successful development of each organism. Something as complex as leg development has to be able to handle variation in each of the components because genetics doesn’t provide precise blueprints of something like bone length; lots of variation will occur between particular organisms even without any genetic variation.
“Let your conversation be always full of grace, seasoned with salt, so that you may know how to answer everyone.” -Colossians 4:6
This is a place for gracious dialogue about science and faith. Please read our FAQ/Guidelines before posting.