As a software developer, I find biology to be fascinating but at the same time extremely complex and counterintuitive. Everything seems to be influencing everything else in subtle ways. You can never separate an organism into distinct modules that do distinct jobs and expose distinct interfaces, like you could a program or a hardware device when reverse engineering. The fact that physics and chemistry are being exploited in most unexpected ways doesn't help either.
But then the thing to remember is that biology is alien technology in the sense that it wasn't created by humans and is still ahead of our current technological progress. It's the only alien technology we've been able to get our hands onto.
Mother Nature only cares about energy efficiency and evolutionary fitness. If that means every variable has to be global, so be it. She is not coding with understandability in mind.
There are very fundamental blocks, like the ribosome. Or enzymes.
Of course over a few million years there was some feature creep, and it turned out that with a little tweaking proteins can be optimized to perform more tasks and then even squeeze out a bit more efficiency out of folding to the task.
You can still expose interfaces, the tricky part is that there are multiple systems you're creating interfaces to -- whether or not you intended it.
Take blood for example. Through a fluid mechanics lens, the role of blood is to maintain a certain amount of flow and pressure in different parts of the body. You can look at it with Poisseuille's Law in mind: F = (ΔP/R) where F is flow rate, ΔP is pressure, and R is (for the purposes of this example) a constant [1]. Low blood pressure = lower flow rate, which means blood spends a longer amount of time in parts of the body than necessary. So blood is depositing oxygen slower than it should. Makes sense that the resulting symptoms would be things related to lack of oxygen -- dizzyness, lethargy, etc. Meanwhile high blood pressure = higher flow rate (which is good), but that pressure is being generated from a fluid pump (your heart), which means your heart is working harder than it needs to so it makes sense that high blood pressure results in heart-related problems. So the role of blood for fluid mechanics is to maintain the right pressure & flow.
Through a biomaterials lens, blood needs to flow easily when it's in the body but not when there's a leak in your pipes (a cut). Its solution is more or less that it recognizes a certain kind of surface as "self" and doesn't bind to that surface, but any other kind of surface attracts plasma which attracts blood.
Through an immunology lens, the cardiovascular system represents a huge vulnerability in the human system because it provides a channel for foreign attackers to reach any of the body's many microservices -- lungs, kidneys, brain, you name it. So it makes sense that the strongest antivirus would be present in the blood stream. That, and the blood is a very convenient transport service for antibodies that need to get to a compromised area quickly, to contain the incident.
Through a biochemistry lens, oxygen is a resource that every cell in the body needs for cell respiration. Blood needs to be able to pick up oxygen from the alveoli in the lungs, and drop it off in a fairly evenly distributed manner throughout the body. It'd be bad if it couldn't even pick up the oxygen (which is why Carbon Monoxide kills you -- it binds to blood and doesn't let go). It'd be bad if picked up the oxygen and dropped it off too early. It'd be bad if it picked up the oxygen and then refused to let go. So the body evolved a really clever system where it uses hemoglobin, which binds oxygen more tightly at higher Ph and more loosely at lower Ph. Conveniently, high concentrations of CO2 lower Ph, so blood naturally lets go of
its Oxygen when it's in areas of high CO2 concentration.
Through a bioelectricity lens, blood is electrically inert. This doesn't get talked about because it's uninteresting, but it is still a design decision that blood doesn't generate or consume electrical energy and does not interfere with the electrical activies of neurons.
You can still draw the interfaces as you would in software engineering. But anything you make will have interfaces to all of the body's systems so you need to make sure you're not throwing off any other equilibriums
I’m probably preaching to the choir, but yup that inherent uncertainty seems to distinguish biology from the other sciences. There is a huge stochasticity & serendipity in everything because there is no intentionality in the design of any biological component; any convergence toward a chemically or physically optimized component or behavior is driven by evolution by natural selection but remains imperfect. Warts, quirks, and all
It also explains how the folks on HN positing “Covid is just the flu” and “long Covid isn’t real” have been so confident yet so gravely mistaken. They are used to other sciences where there is much less room for uncertainty
Yep. For better or worse i've been up to my eyeballs in "hey, you know, biological systems are actually insane!" since high school, and now through a few advanced degrees :)
I've been working in tech more than normal for a few years now, and the endemic nature of "the andy grove fallacy" among my coworkers has ceased to startle me, but it's just kind of bothersome every time.
Despite the common trope that software/hardware engineers think they understand everything, automation can make exponential development possible in bio. I know someone who works at a biotech company, and the stories I hear lead me to believe that traditional thinking in e.g. medicine and insurance, and from colleagues, really is a bottleneck. Maybe exponential growth will never apply to the number of diseases cured, but it can and should apply to the hardware and software that facilitates the next iteration of biological discovery.
It's not that software/hardware engineers think they understand everything, it's that the absolutely common idea "my way of thinking about understanding things is appropriate in this other domain" breaks down very fast.
I'm not speaking about running hospitals or dancing with insurance companies. I'm speaking about much earlier in the pipeline: fundamental, blue-sky biological research is fundamentally different from software. The reason is that we are not studying designed systems; the effect is that there is so much that we don't know that we don't know that things that sound easy are in fact basically impossible because the prior knowledge is simply not established. (Until they aren't. When does that change? First slowly, then all at once.)
As a biomedical engineer, i'm on board with hardware and software improvements: it's kind of my job. The trick is knowing what you're doing, what you aren't, and what you can expect, and to balance confidence in the value of what you _can_ tightly constrain and design versus humility in accepting that the natural world simply doesn't care.
I think what bugs me the most is the common assumption that we just need better modeling tools to make a lot of the messy lab work go away. Everyone who raises this idea seems to think it's original and revolutionary and a no-brainer, but there's at least 40 years of bitter experience in biotech and pharma development saying otherwise. Even genuinely impressive feats like AlphaFold get inflated wildly in importance (and used to retroactively bash experimentalists for all the time they wasted by not listening to software people). Actually speeding up the entire process of biomedical research requires improvements across the board in many different fields, and it's not something that is magically going to be solved by computer science wizardry alone.
I primarily do modeling, and I am first in line to tell anyone who is willing to listen that the very first thing you should do to learn to be a good modeler is spend a year at the bench first in the problem area you want to model.
Wet lab is so informationally dense in terms of personal knowledge that it's not even funny.
Natural selection may not select for simplicity. It may not try to reduce dependencies. A gene can depend on thousands of other genes. And a gene can carry out multiple functions
Natural selection doesn't "select" for anything except the survival and continuation of species. Everything else that it tends to favor is merely a correlation with that - so you have to expect a certain level of variation and exception.
This one's pretty straightforward; natural selection has two very common approaches to dependencies:
1. If the presence of something is not a reliable feature of the environment, dependence on that something will be ruthlessly weeded out.
2. If the presence of something is a reliable feature of the environment, dependence on that something will be created and grow to permeate most existing systems.
The first effect has a lot of popular awareness; the second doesn't. They're both important.
Counterargument to 2: the air is largely nitrogen, but humans do not utilize it during respiration. Thus, this reliable feature of the environment has not been selected as a dependency.
I'm not necessarily arguing it's untrue, just not quite the hard-and-fast rule that 1 is.
> I'm not necessarily arguing it's untrue, just not quite the hard-and-fast rule that 1 is.
This is pretty subjective. If you grow up with sufficient dietary iodine, that's worth 10-15 IQ points (a gigantic amount) compared to the counterfactual. Iodine supplementation is highly effective in populations that have presumably been going without for, at the very least, many hundreds of years. This is the reason we iodize salt.
With that in mind, how much of a hard-and-fast rule is #1?
The other side of this argument is that dependencies can be surprisingly subtle. This is pure speculation, but the thing that leaps out at me about atmospheric nitrogen is that oxygen is extremely corrosive. If the atmosphere were more oxygen and less nitrogen, lungs of today might experience a lot more wear and tear than they are currently designed for. (Or not! I really have no idea. But I think the hypothetical has some thought-experiment value anyway.)
Yes indeed, there are examples of dependencies that have been removed by evolution.
For example cobalt and nickel were both essential for the first living beings. However, on the continents they are much less accessible than in the oceans.
Because of that, most terrestrial plants have lost the dependency on cobalt (which is why they cannot provide vitamin B12 to those who eat them) and most have also lost the dependency on nickel (except that for many terrestrial plants nickel, while no longer strictly required, can still be useful by enabling them to use urea as a source of nitrogen, when other sources of nitrogen are scarce).
> 1. If the presence of something is not a reliable feature of the environment, dependence on that something will be ruthlessly weeded out.
Something is only weeded out if it is sufficiently disadvantageous. There are lots of examples in biology where mutations are neutral and allowed to accumulate. There is also a lot of junk DNA, and even though we've found that some junk DNA was misclassified, there still is a lot of junk DNA in each genome
If a reliable feature is evolutionary advantageous, then new systems will come about to take advantage of that (and may out-compete old systems, indeed, to the point of extinction), I agree.
But, if a reliable feature isn't evolutionary advantageous, it can be ignored or even removed.
Negative features may need to be managed, rather than taken advantage of. Examples: ionizing radiation, landfill.
In mammals, the dependance on a certain body temperature.
The fact that it is regulated to a margin of 0.3% shows quite how important it is. Increase or decrease the body temperature by 5% (on an absolute scale, so Kelvin), and you will be a dead human...
> The fact that it is regulated to a margin of 0.3% shows quite how important it is.
This is an excellent point. It can be applied to a huge number of biological phenomena, but it almost never is.
An example where this idea actually has made it into the mainstream is body weight. I read an article that pointed out that, yes, we're fatter than we used to be, but regulating body weight by measuring energy intake and expenditure is doomed to fail. Articles making this claim are a very rich genre, but I found the particular argument fascinating:
People routinely maintain the same body weight, within say 5 pounds, for years at a time. This is a level of accuracy that we do not have the technology to achieve by using measurements of energy inflow and outflow -- we can't measure dietary energy that precisely! (And modern technology is, by most standards, very precise.)
Thus, there is a very strong suggestion, based on this incredible stability, that your body is regulating itself to a certain weight. If you adjust energy intake, or outflow, other things will adjust accordingly; those are unlikely to be primary determinants of the system's behavior.
A question from a layman: Could changing body's temperature by a few percent, for a small amount of time, be used to heal from some diseases, like cancer or diabetes?
Some diseases, yes; people do this all the time and it is called "running a fever". Fevers are not induced by the disease; they are a strategy you use to kill the disease.
Cancer or diabetes, no. Diabetes is not even a "disease" in the sense of an organism that can be alive or dead.
it's just a slightly more extreme form of the glove-and-hand principle (versus lock and key). Of course even this turns out to be a spectrum. there are enzymes/protein that are more lock-and-key, and proteins that interact very heavily. One extreme is insulin receptor (which totally mangles insulin out of shape on binding). IIRC, another interesting one is antibody evolution, as B cells mature they go from producing antibodies that are gloves-in-hand to antibodies that are lock-and-key, as measurable using IR spectroscopy: https://www.pnas.org/content/103/37/13722.short
I don't like the headline. As if the protein decides to do so. Maybe this sounds a little bit less "active" but probably also less sensational: "Some Proteins Perform Different Jobs By Changing Their Folds"
One of my pet peeves is science journalism abusing the active voice, often in ways that almost anthropomorphize the process. The absolute worst are phrases like "Nature designed this protein to...", which I have seen otherwise reasonable, rational writers toss about carelessly. (It's like the inverse of the passive voice used to describe police shootings: "the suspect was hit by gunfire", etc.)
Well, you can't reason with unreasonable people :) That's the actual problem. Also, I strongly disagree with your view that a lot of them are reading scientific articles. The root problem is lack of science education/curiosity which needs to be addressed specifically, rather than forcing the rest of the world to adopt certain phrasing.
Personally I like the romanticism/poetic style, it makes articles seem less bland, which makes me want to read them even more.
