Thanks for that. I found it to be quite fascinating but at the same time somewhat unsettling. The depth of complexity seemed limitless. As though there would be no end to the layers of abstraction one could reason about - and must reason about, if we are to truly understand ourselves.
I think finding it unsettling is actually a good reaction. I think people get too encultured into simplistic views of the world. It's good to recognize how fundamentally deep and strange the universe is with respect to our everyday experience of it, if for nothing more than the sake of perspective.
I've spent a great deal of time wondering why biological systems are so complex. Seeing this videos leads me to ponder this question yet again. I keep coming back to the same answer. Evolution as a process proceeds forward without regard for complexity. In other words, the outcome of an Evolutionary does not have to be simple - it just has to work. I fear that most biological systems are in many ways unnecessarily complicated, which makes research in Biology and Medicine even more difficult.
Well there's all sorts of issues. But what it essentially comes down to is that evolution works in tiny steps, not in great leaps. When "trying to achieve" goal X, adding a tiny bit of complexity is simply much more likely than a simplifying "refactoring".
And then in terms of the biology itself, you will run into things that we might think of as "complications" but which are actually for the sake of redundancy/fault tolerance, etc.
I remember watching a series of biology videos where a particular messaging pathway was explained. It had at least 5 or 6 proteins. But then the lecturer described that by having so many proteins in the chain, the messaging is exponential. So what initially might seem like a complex way of sending messages is actually an amplifier!
Nature does not solve problems. Molecules bump into each others. There's no purpose in their doing so. There is not simple for nature and complex for us. If you want a scientific approach to complexity, you would not insert a term in the definition that reads as 'us', or 'nature', or frogs.
Alright, there is no inherent purpose (something I always say to others and here I am giving the opposite impression :) but if we agree that things 'are' because they sustain some abstract structure in some context then can't we allow saying 'to sustain' is a problem to solve ? and if the context is very diverse thus high in complexity. And as we lack information, the observed structure will elude our understanding, 'complex for us'. From my very limited point of view, I'd believe nature follow some sort of simplest path possible (~lagrangian) law, hence my 'simple for nature' expression.
Living organisms are not structures. Not sure if that's what you say, but I have heard this many times, in the context of origin of life. Where life is seen as self-replicating structures.
The complexity of a geometrical structure (like a cube, or a protein) is one thing. The complexity of a process that fulfills a function (engine of a car) is another thing. One is static, the other is dynamic. One just sits there. Like a pile of rocks. The other one takes an input and process it into an output, in a specified, repeatable manner.
A pile of rocks, in a particular context, solves the problem of surviving the surrounding conditions. Is it the simplest solution? Who cares!
DNA replication on the other hand? What did Mother Nature drink to come up with this problem and try to solve it?
That's why I added 'abstract' it's not fixed, things move underneath but the abstract properties remain.
Funny about the rock thing, I never know what to make about them, I would need to add an 'adaptability' rule to my core ones.
DNA replication did not emerge from the early organisms. Before replication, we could investigate about DNA or prototypical forms of DNA as forms of variable input transformation. This kind of meta-encoding (sorry programmer POV leaking) of input tranformation is probably not the first minimum but pressure + time pushed the versatility needs up (again programmer POV).
I'm a programmer too, but also a physicist, and it's also hard for me to read through explanations/narratives that use a lot of metaphorical language. And there's a lot of that in evolutionary narratives.
That's really a very profound way to see it! Thanks.
I'd add to that: there is even a chance (quite high btw) we are not yet seeing all the problems nature solves AND maybe we never will even though we are constantly and exponentially growing our understanding... always, forever, now.
Evolution DOES proceed with a lot of regard for complexity. Evolution is not a person. The processes by which the evolution is supposed to go about are very much complex.
Complexity is something that can be clearly defined in terms of probabilities and hence measurable. If it's stated that these complex (made of {large number of} parts) molecules _work together_ with these others complex (also very large) molecules, than this all process, in order for it to fulfill a function, requires a set of very specific values for very specific parameters.
This is measurable complexity. And if one says that in time evolution will solve a problem, you can make the liaison with the complexity of the process through the probability of its occurrence.
This is what makes the concept of junk DNA [1] non-surprising [2].
Often complexity would come at a cost though, and evolution would automatically end up optimizing for it. Such costs may include energy costs, vulnerability to faults/diseases, etc. With such elaborate mechanisms for DNA wrapping, replication, etc., I wonder how much cost junk DNA brings to the system (vs. benefits in terms of fault-tolerance and evolution).
Evolution proceeds without regard for complexity, efficiency, or anything else for that matter, except survival. I think the most common misconception about evolution is that it is not a system of "best" but rather "just good enough".
For example, do you like your lungs? They work pretty well, right? But how long can you hold your breath under water? Obviously cetaceans have much better lungs that humans, but that's because our lungs are just good enough for running around on a savannah. If we were forced to dive to incredible depths to search for our food, we'd need to evolve better lungs, as the cetaceans have done.
Before the dolphins get to smug, though, they don't have the best lungs either. The title of best lungs goes to the birds. Whereas all other air breathing animals have bidirectional flow in their lungs, birds have evolved a system of continuous unidirectional flow. Turns out that bidirectional flow is not good enough at high altitudes with an extremely high metabolic rate.
As for what makes Biology and Medicine difficult, I think the answer is: time. That is, aside from dealing with systems composed of more moving parts than physics or chemistry, biology is also less open to direct manipulation and experiments take, on average, much longer for the amount of information gained. Eventually, though, I don't doubt that we will understand biology as well as we understand the other two. At the moment we're in the "alchemy" stage of biological research. We have a sense that there are units of action and rules governing these units, but we don't know what those units are or how they interact. A rock you dig up from the ground is, also, a rather complex system difficult to understand until you have knowledge of chemicals, molecules, bonding, etc. We just need to figure out what the "atoms" of biology are.
I believe the exact opposite: Biological systems are extremely optimized and add complexity only when needed.
Of course, in order to proceed in a meaningful way this conversation we should first establish what exactly complexity and optimization means, in this context.
So if you feel like it, please share a specific example.
The mere presence of Junk DNA represents a potential layer of unnecessary complexity. This makes the job of deciphering the function of each gene just a little harder. While there is much debate about how much of our own DNA is junk, there is little doubt that a considerable amount of it is just junk.
There are also cases where molecular machinery in our cell is repuprosed to carry out novel tasks unrelated to its design. The first example that comes to mind are the voltage-gated potassium channels. These proteins are usually involved in the electrical signalling of the nervous system. Why the voltage-gated potassium channel Kv1.3 functions as part of the immune response in T-cells makes no sense to me. Yet this voltage-gate potassium channel is an indispensable part of the cascade of events underlying the immune response. Another example of a protein with numerous and possibly unrelated functions is p53.
And then there is of course the example of the laryngeal nerve. This nerve makes an unnecessary loop down the neck and then back up to the larynx. This odd quirk is especially exaggerated in the giraffe. Why this is so makes no sense.
The best explanation I have to such weird design quirks, which exaggerate the complexity of biological systems, is that Evolution as a process has no regard for complexity.
Why the laryngeal nerve is the way it is makes much more sense if you look at is as a series of trillions and trillions of little tiny steps.
At some point a very long time ago vertebrates had very different designs from the contemporaries. As they evolved from strange filter feeders in to jawed creatures, this nerve extended itself longer and longer as creatures with a neck came in to existence. The important thing to see here is the lengthening of this nerve didn't cause selection pressure against it. If any vertebrate animals did develop other, more efficient, nerve routing, it did not give them an advantage to survive unto this day. None of the things we are calling quirks were quirks far back in the past. They have been 'stretched' by billions of iterations. It would be complex, by all definitions of the term, if not impossible to put DNA and all the cellular functions in any semblance of order that would make sense to a person.
If evolution has regards for anything, it is survival. No other metric, be it efficiency or complexity, matters unless you produce offspring that produce offspring.
Yesterday I stumbled across a quote from Francis Crick which enhances your view actually: God is a hacker not an engineer.
But since there is a huge lack of understanding at this point in time about the DNA junk sections, the p53 protein, the formation of the protein's quaternary structure, the K channels it's hard to accept an answer as definitive.
A fully functioning life-form depends on so many things going right. With such precision needing to be met, I'm just surprised that there aren't more debilitating genetic diseases out there and that so many people are seemingly healthy.
Of course, it also has plan B and C in case plan A fail and is remarkable at adapting. Just for example, look at all the people with genetic diseases still hopping along day after day.
The WEHI videos are really fantastic, because they do a good job of 'random-walking' components into place. Compare and contrast the WEHI videos to the Harvard molecular visualization videos, where pieces come together with unrealistic symmetry:
This unrealistic symmetry and lego like snappiness, plus: in almost every visualization of these tiny tiny structures and processes involving them most of the "environment" is missing!
Think about all the school, high-school and even college visualizations showing the organelles f.e. just hanging there pinned into place. Between them wast areas of ... what?
Now look around you. What do you see? Maybe the furniture of you room resp. office... now wave your hand in front of your eyes (nooo, they are not laughing AT you ;) ... did you feel the interaction of the molecules "in the air" with those of your skin? They are filling the gap between us and the things we are capable to perceive with our eyes...
The "invisible" content of these gaps is interacting with the visible parts. At molecular levels those invisible elements contribute hugely to the non-deterministic, non-lego like snappiness, to the "random walk" way of things working out.
This kind of insight I think should be replacing the "deterministic machinery" view when describing and picturing molecular and micro-biological processes.
Visualizations by Dr. David S. Goodsell present this kind of ideas in a fascinating way.
I don't think the 'missing environment' is as much a big deal. Without getting too reductionist: You have to isolate the components that you're showing otherwise it's just a big mess that's hard to understand. I think the implicit "there's actually a lot of stuff in between there" is not a big deal, much as the implicit "these molecules aren't really colored" is not a big deal in the service of explanatory power.
The reductionist view is not complete, but it's not wrong. Hypersymmetric assembly and molecules that 'know where to go' is wrong, and in a very subtle way that biases perception. The guy who makes the harvard video, in his TED talk, goes on and on about beauty in science. Well yeah, he created that beauty himself. The real system is sloppy and kludgey.... which I suppose could be 'beauty' in its own way, but very different from the way it's presented.
>>You have to isolate the components that you're showing otherwise it's just a big mess that's hard to understand.
That's for sure. No way to depict all of it at once...
>>I think the implicit "there's actually a lot of stuff in between there" is not a big deal, much as the implicit "these molecules aren't really colored" is not a big deal in the service of explanatory power.
I disagree partly here. I imagine (hope! ;) "professionals" do know about the missing and false color "stuff" BUT every picture flushed towards broad public (vs professional publication f.e.) consumption should contain as much context as possible, context "disturbing" the mechanical picture presented of bio-molecular actually all natural reactions today. "knowing where to go" is wrong as you say and it would not "fit" / "work" / "be accepted" / "look believable" the moment the surrounding soup gets visible.
Leaving this kind of context out, i think, propels many misunderstandings about the way science is actually understanding nature.
>> which I suppose could be 'beauty' in its own way, but very different from the way it's presented.
This is actually a heavily criticized animation within the computational biophysics community. Our intuition for physical determinism breaks down at the nanoscale. One major fault of this video is that, at low Reynolds number, there's essentially no inertia. So the representation of kinesin taking deterministic, 100% processive steps is an absurd idealization. The motion is dominated by thermal fluctuations (which, by the way, are occurring on a nanosecond timescale).
At the same, even as a biophysicist, I think these animations are superb learning tools. Those beautifully rendered animations actually describe the mechanism in far greater detail than, say, textbook diagrams. One should just be aware that fluctuations dominate at these scales.
I remember watching these animations in high school biology class. They've definitely had this big long-term impact on how I visualize and reason about cellular and subcellular components. Just always see little components bobbling around and randomly flipflopping.
Growing up in an environment that was in certain regards ignorant, even hostile against science and technology I remember hearing about DNA for the first time I had this uneasy feeling that DNA is pretty much like computer code.
Not unit tests, but a fair bit of in-line verification that the correct base has been added. There are numerous DNA proofreading mechanisms to prevent errors. If they didn't exist, we would be riddled with mutations.
More specifically, it's DNA-Polymerase who proofreads. You should also check out the T-Cells (innate and acquired immunity) work. It blew my mind the level of complexity, diversification, etc: There are cells that perform profiling, cells the memorize information about viruses (that's how vaccines work) already encountered, etc.
Molecular biology is like entering the rabbit whole.
Imagine a wormhole opening up in front of you and a computer from the unimaginably distant future drops out in front of you.
You very likely wouldn't know it's even a computer. Countless billions of hours a futures mankinds time has been poured into it shrinking each component to the atomic scale. Generation after generation of hardening against future hackers attacks, and even self aware viruses have increased the machines complexity beyond human understanding. The code the computer interprets would be a lost language from a civilization that doesn't exist yet.
And that is pretty much where we are in understanding the interworkings of DNA today. Nature has a 3 billion year jump on us, and it's going to take us a long time to unravel it all.
