Skip to main content

Biological News

Engineering Biology, For Real?

Any article titled “How to Engineer Biology” is going to get a look from me – and when I’m referenced in the opening paragraphs, especially so. This is a piece by Vijay Pande in Scientific American, and I get called out for my naming of the “Andy Grove Fallacy” (found in this post and the links therein). That’s the idea that the drug industry makes slower progress than Silicon Valley does, therefore applying the engineering and management styles of the Valley to drug industry will speed things up. I’ll spare you the suspense: that’s wrong. See that last link above for why.

Pande is here with the news that whatever merit my point of view might have had, the Valley is now here to melt those objections away. There are certainly parts of his article that I agree with, such as the idea that an engineering approach to biology is more easily seen (and accomplished) in the engineering of the “tools we use to manage biology”. One of the striking things about molecular biology (indeed, a striking thing that makes it possible as a separate science) is how those evolutionarily-developed tools can be adapted to new purposes by humans. The polymerase chain reaction, DNA endonucleases, homologous recombination, DNA ligases – the list goes on for quite a ways past that. We’re able to take these processes and the enzymes that accomplish them and put them to uses both inside and outside the cell. We most certainly can engineer proteins (sometimes), and we can engineer enzyme function (sometimes) – neither of those are exactly straightforward tasks. I think that it’s correct to say that learning to borrow and modify such things really is a triumph of human science and human engineering, and that it’s nowhere near being exhausted yet.

But I’m afraid that that’s a minor point in the article, and I have problems with most of the major ones. To my reading, the piece takes a number of very broad leaps, disguised as dance steps. For example, after talking about the modularity of biology (building blocks such as nucleotides and amino acids being assembled into biomolecules, cells being assembled into organs), this statement appears. “Once we identify the Legos in biology and their properties, we can engineer them and even mix and match them to design novel functionality“. And that’s true, but when it’s put like that, there’s some natural confusion about how many of these “Legos” there might be. We’ve identified a few, but there have been some profound surprises along the way. To pick one, how about the number of different kinds of small RNA species running around in the cell? There’s a whole set of Legos, a whole set of sets, and we had no idea until fairly recently that they were in there.

Even if there were a relatively small defined number, this statement would have problems. There are around 80 stable chemical elements – would it be appropriate to say that “now that we’ve identified the Legos in chemistry, that we can engineer them and even mix and match them to design novel functionality”? Well. . .yeah, I supposed so, but that skips over a few million details and makes all of chemistry sound like a fairly straightforward exercise. Biology, I should add, is far worse.

This is one of the points where the straight engineering viewpoint (as in the famous “Can A Biologist Fix a Radio” essay) breaks down. Radios (of the kind discussed in that article) are human-built objects made of discrete parts. You can see all of them clearly – even if you don’t know what they do (yet) you know that they’re there. But that’s not the case in a cell. We had no idea about small hairpin RNAs, double-stranded RNAs, microRNAs, circular RNAs and all the rest of the menagerie. Didn’t know that they were there. Didn’t see the functions that they were performing, didn’t know that there were such functions. And we didn’t see the workings of the cell and say “Ya know, there’s gotta be a bunch of small regulatory RNA molecules in there, that’s the only way this thing makes sense”. Not a bit of it.

So when you casually say “Once we identify the Legos in biology” you’re actually asking for a great deal, and by disguising it in terms of similarly-sized little building blocks, you actually are confusing the issue. Lets say that the Lego blocks in this case are the five major nucleotides in DNA and RNA. We’ve identified them. Does that mean that we understand their systems well enough to mix and match them? Well, crudely, yes – we can go in and change a genomic sequence. But do we know what happens when we do that, and why? Not so often, not at all. Can we add in the novel functionality we want? Sometimes yes, sometimes no, and it usually takes empirical tests to see what the answer is going to be (and even when we get that answer, we often don’t know why we got it, as with protein expression experiments). Now that we’ve identified the five common nucleobases, does that tell us about the weird little uncommon ones, and why they’re used? It does not. Does it suggest all the strange RNAs mentioned above? Not at all. Does it tell us about the promoters, stop codons, histones, transcription factors, RNA polymerase enzymes, ribosome entry sites and all the other wildly complex things that go into the expression of the gene whose sequence we’ve just edited? Not a bit of it. “Identifying the Legos” only takes you so far. It’s necessary – vital – but not sufficient.

That Lego stuff was Pande’s “Principle 1”. The second principle is “Repeatability and Reproducibility”. His point is that biology has been held back by the tricky, delicate, bespoke nature of many of its experiments, and that modern equipment allows things to be run in a much more reproducible way. Like the Lego section, this one is true, but only so far. Here we go:

The identification of biomarkers (chemical substances we can measure and then target) for disease is currently driven by discovery via a bespoke, one-off process—so the discovery of PSA for prostate cancer, for instance, does not suggest a biomarker for ovarian cancer. Introducing machine learning into the process, however, can turn this handcrafting into assembly-line production.

This is Lego-istic thinking again: the belief that these things are discrete graspable units that can be handled as if they were lines of code, hardware components, or indeed Lego bricks. But we’re not to that point yet. I’m not sure how to put this gently, but the discovery of PSA may not even suggest a biomarker for prostate cancer itself, for starters. It does in some patients, but not in others, and whether it’s of benefit in the general population is very much a subject of debate among physicians. You don’t get this so much in engineering, because engineering is so much simpler. Saying “a biomarker for ovarian cancer” makes it sound like there is a biomarker out there, waiting for machine learning to uncover its benefits for a disease called ovarian cancer. But Pande himself is far too well versed in this stuff to really believe that. “Ovarian cancer” is not one disease with one biomarker – like almost all cancers, it varies from patient to patient, and over time in any individual patient, and from cell to cell inside any individual tumor in any one patient. Bridges do not work this way, to use an engineering metaphor that appears in the article, because bolts are bolts. But not in the biology of human disease.

The article would have you believe that there are machine learning companies that are right now churning out biomarkers for all sorts of diseases, and that “they can now mass-produce many tests in a predictable, precise and repeatable manner”. My main response to this is to ask for the names of some of these companies and the tests that they have actually brought to the market.

Pande’s “Principle 3” is “Testing and Process Engineering”. He says (absolutely correctly) that in engineering “the need for testing is obvious, how to test and what metrics to measure success are not. So, the choice and engineering of key performance indicators (KPIs) is critically important here; without this guiding compass, a project could go in the wrong direction.” But then the argument is that biology and biologists have missed out on the application of these KPIs to their own field. Pande goes on to say: “Now, a new wave of bio startups—drawing on engineering and computer science—are identifying KPIs for measuring molecules synthesized to protein expression, numbers of cells screened, and much more“. I think that something must have gotten turned around in the editing there, unless there should be a comma after “synthesized” and the following “to” should be struck out. But again, it would be useful to hear the names of some of these new wavers, and why they think that some of these are key performance indicators. When they “measure molecules synthesized”, what are they measuring, exactly?

Moving on, “Principle 4” Is “Borrowing From Other Disciplines”. Pande says “. . .the rise of numerous, novel quantitative measurements of biology—i.e., big data sets in biology—has opened the door to incorporating other engineering approaches“, and while that’s true up to a point, I’m not so sure about his take on this. For example, he says that “By applying the materials-science based engineering technology he learned in solar cell materials design to food, James Rogers used techniques from nanoscience to create nanoscopic barriers that protect fruits and vegetables from spoilage.” As a correspondent noted to me over the weekend, this makes it sound as if these “nanoscopic barriers” were being designed layer by layer under a scanning tunneling microscope or something, when what they actually are, are hydroxy-fatty acids and their glycerides to make a coating that goes on more uniformly. What’s more, it’s based on what we know about plant-based waxes such as cutin and suberin – and is in fact produced from them. This is a perfectly good invention, and looks to be really useful, but does not herald the advent of solar-cell engineering techniques applied to living systems.

Finally, we hit “Principle 5”, which is “Reinventing the Process Itself”. That title gives me flashbacks to various HR initiatives I’ve been roped into over the years, but getting past that, what he’s saying is this:

“The challenge in biology lies in breaking down the problem into steps and often reinventing the process itself. But once the desire to consistently improve performance (what (Andy) Grove was suggesting in the first place) moves biology from bespoke, artisanal approaches to designed, scalable processes, even seemingly modest performance increases can make a difference”

OK now. I know exactly what’s being said here – my response is not due to incomprehension. But this exemplifies what to me is a problem with the entire article. It references a number of engineering practices, asserts that it’s now possible to do these things in biology, and simultaneously gives the impression that no biologists had ever thought of trying any of them before. And there are major problem with both of those. How, for example, does anyone think that the Structural Genomics Consortium has been running through so many automated protein crystallography and X-ray structure experiments over the years? How did PCR get optimized to the workhorse procedure it is now? How did monoclonal antibodies move into industrial production? How, indeed, did something like Sanger sequencing get developed back in the early days? By breaking the problems down into pieces and optimizing them separately.

The objection to that might be that those are examples of what Pande was talking about earlier, the application of engineering to the tools of the trade. But my response would be that (1) this shows that such approaches have been going on for a long time (as applied to tools and processes) and are not some new revolution, and (2) that, on the other hand, such an engineering mindset is still not possible for the basic-research side of the business. Pande’s article really tap-dances around that latter point. Engineering just somehow is going to do these things, and it’s on to the next bullet point.

Perhaps the last sentence of the article is where the problem gets stated most clearly:

The question now isn’t whether this is possible in biology or not, as the Grove fallacy argued, but how to do it, given where we are in engineering biology today.

I think we’re dealing with a fundamental misunderstanding here. When I’ve written about the Andy Grove Fallacy, I have not been suggesting that it’s impossible to do biology in some sort of organized fashion. What I’ve been emphasizing is, though, that the things that make such an approach so productive in Silicon Valley will keep it from having the same effect on drug research, at least for a good long time to come. The challenges in hardware and software design, though significant, yield much more easily to human pressures than those of biology and medicine, and twenty paragraphs of repeated assertion to the contrary doesn’t change that much.

Consider the nematode. It’s a terrific little animal to study; Sydney Brenner was rightC. elegans has a limited number of somatic cells (959 in one sex, 1031 in the other) and we know the exact lineage of every one of them through systematic study. It has around 20,000 protein-coding genes, and it has of course been completely sequenced in great detail. Our current technologies allow us to step in and mess with those genes individually. That has, in fact, been done (and far more than once, using different technologies).  The nematode proteome has been studied in great detail, under many different conditions (stress, age, mutations). And so on.

My point is that if the nematode were a product of Silicon Valley, we would have more than enough information in hand by now to build one, reverse-engineering it like a new device or a pile of source code. But we can do no such thing. The nematode has been subjected to systematic, engineering-driven analysis fit to to Vijay Pande proud, but we cannot assemble a single one of those thousand cells. There are unknown things going on inside each of them that will win people fame, fortune, and Nobel prizes once we figure them out, of that I am absolutely certain. Real functioning nanotechnology is at this very moment rolling along a nematode’s genome, spitting out messenger RNA that in turn is being ratcheted through ribosomes in ways that we’re still in awe of (and still figuring out the details of, for that matter). Nothing of the sort is happening in an iPhone, and I don’t mean protein biology, I mean extremely important things that we don’t understand and may well not even suspect the existence of. That’s because we built the iPhone, in every detail, and we found cells waiting for us with a three-billion-year head start.

So rather than refuting or superseding the Andy Grove Fallacy, from my viewpoint Pande’s article gets a running start and takes a cannonballing high dive directly into it. I will be very glad to hear opinions on that. . .

Here’s the Wavefunction take on the article (“Whatever the complexities of challenging engineering projects like building rockets or bridges, they are still highly predictable compared to the effects of engineering biology”), and here’s Keith Robison’s (“What Pande is far too optimistic about is the difficulty in figuring that out, particularly when trying to deliver therapies”)

59 comments on “Engineering Biology, For Real?”

  1. SP123 says:

    What surprises me is that the SV set doesn’t just declare victory and say that computer engineering principles have transformed biology, because you could make a reasonable argument for that.
    Think about the things we can do- you can go to a website and order mass-produced enzymes to cut useful places in DNA to let you move biological elements around however you want. There are factories with machines that will churn out strands of DNA that you can reassemble into useful biological objects- we can pretty easily make a library with every mutation at every position in a given protein to investigate how it works. Then there are factories of different machines that will take DNA and churn out billions of bases of sequence data per hour. Compare that to the fact that almost 50 years ago the expression and characterization of EcoR1 was a significant publication, or to the well-known cost of the first draft human genome.
    Without the entire industry that’s grown up around molecular biology, most of which relies on principles of engineering and manufacturing, we’d be completely helpless understanding any of the things we do understand. So the SV people should just say that biology has already become a subdiscipline of engineering, and that their approaches have gotten us to where we are today- because biology is so vastly more complex, we’d have no hope if we hadn’t learned from engineering.
    (Of course the reason they don’t do this is that most diseases have not been cured despite the advances in biology, so there’s fame and fortune out there for whoever can take it further- not realizing, as you note, that there might not actually be a solution for every biological problem we’d like to solve.)

    1. zero says:

      That position would be warranted and relatively accurate, but also uncontroversial and with no room for the promise of future disruption.
      In other words, the truth in this case has very little VC investment potential and so will get very little daylight from one side of the debate.

  2. Hap says:

    Doesn’t the Lego analogy assume a modularity and limitation to interactions that isn’t in evidence? Lego blocks can only interact in certain well-defined ways and their numbers and structures are known. We don’t know all the pieces in biology, nor how they all fit together, or the ways in which they interact. How can you assume an overall design if you don’t know how you can assemble the pieces?

    This sounds like Dunning-Kruger rearing its ugly head again.

    1. mjs says:

      What about the intrinsically disordered Legos?

    2. eub says:

      Right, the Legos are utterly assuming the conclusion.

      How about Silicon Valley shows us how to apply engineering principles to software, as an easy start? Much simpler stuff, made by humans instead of by billions of years of evolutionary stewing. And “software engineering” is a crock, because even software is a far cry from Legos.

      Our success in software is built on brilliant computer science research by painstaking artisanship, while software engineering waves its hands over it. (Seriously, imagine programming with only data structures and algorithms that were known in 1960. Lots of good work done by then. But so much not!)

  3. Mad Chemist says:

    Looking at Pande’s bio, he really should know better than to say the things he’s said in the article.

    1. Magrinho says:

      Of course he knows better. But check his most recent job title.

      Whose Law is “Follow the money”?

      1. Passerby says:

        Pande is a smart guy, but in addition to working at a VC firm whose interests are in funding AI and technology-backed biology companies that would benefit from the kind of thinking he is describing in his piece, he has also spent most of his career working in reductionist molecular dynamics simulations and distributed computing, fields which have a very tenuous relationship with emergent biology and drug discovery at best. I am not saying the choice of fields is his fault, but one does have to wonder if this background makes him see the messier parts of biology through what Derek once called “Silicon Valley Sunglasses”.

        1. Dionysius Rex says:

          Maybe the Silicon Valley sunglasses have a special type of polarizing filter that only allows one to see things in the desired dimensions?

      2. GM says:

        There is some debate as to who actually said it first, but it is often attributed to Associate FBI Director W. Mark Felt, who, 30 years after the Watergate break-in finally revealed he was Woodward’s* source “Deep Throat”.

        *No, the other Woodward.

        1. Derek Lowe says:

          To the best of my knowledge, that attribution to Felt is an artifact of the screenplay to the movie “All the President’s Men”, interestingly:

  4. Looking at resistance... says:

    Hmmm… We know all the stable elements. We have working DFT and virtually infinite computer power. Where the %!*&^%#! are the room-temperature superconductors. Way less complex problem that understand cellular biology…

    1. cthulhu says:

      Physics Nobel laureate Robert Laughlin has a good joke about this kind of thing in his fascinating non-reductionist popular science book “A Different Universe: Reinventing Physics From the Bottom Down”; goes something like this:

      A small country has a revolution, and the new government is executing its new plan in literal fashion. Two higher-ups in one department are being asked for their last requests prior to the firing squad. The first says, “Before I was in government, I was a theoretical physicist. My last request is that you assemble a conference of our country’s physicists so that I can present my theory of high temperature superconductivity.” The second one then says, “I am also a physicist. My request is that you shoot me before he gives his lecture.”

  5. Emjeff says:

    I have an analogy for biological research that I give to layman, and I’d like to state it here. Say you are in your living room, and you want to see what is going on upstairs in your bedroom. You cannot go upstairs and observe directly, but you do have a bunch of mirrors that you can set up cleverly so you can see into you bedroom. Ah, but some of the mirrors are cracked, others are distorted, and in any event, you will need to correct the inevitable distortion caused by that many mirrors. After weeks of testing you might even think you’ve observed something in the mirrors that later would be found to be completely wrong, thus necessitating a need to re-set the mirrors.

    This is biological research. Everything we observe is observed indirectly, and most of it is distorted in some way, which may cause us to make erroneous conclusions(see sirtulins). In any event, it is far harder than writing code, and those who are as arrogant as Pande will so learn just how hard it is.

    1. secret sauce says:

      You forgot to add smoke with the mirrors – a key component for merchants of science. Pande may be arrogant, but he’s certainly not ignorant – he’s an expert player, both in academia and business.

  6. Peter Kenny says:

    While I’m sure that we’ll be able to design some better components, cells are dynamic entities and their components need to be created (and destroyed) according to the varying needs of the organism. With respect to measurement, it’s worth pointing out that unbound intracellular concentration is not something that can currently be measured in vivo for drugs (or other physiologically-relevant compounds).

    1. Derek Lowe says:

      That’s perhaps the central problem of a quantitative, data-driven approach to compound binding: we really don’t know where the drug candidates are in the cell, what they’re binding to, and what the equilibria are like.

      1. tlp says:

        and those equilibria are never really reached

        1. Hap says:

          Is equilibrium ever really reached?

          1. MrRogers says:

            It’s called death

          2. Dr CNS says:

            Not really. After death we decompose by chemical reactions

            As for unbound drug concentrations, remember not all ligand/receptor systems follow the free drug hypothesis.

  7. MM says:

    “…because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns – the ones we don’t know we don’t know.” – Donald Rumsfeld. Might as well have been referring to biology.

    1. SP123 says:

      I was always disappointed he left out the fourth quadrant, the unknown knowns, things we don’t know that we know, because that has a lot of application in large drug discovery organizations. The data is out there somewhere in the organization but it’s so siloed or disorganized that nobody knows how to find the information they need.

      1. ScientistSailor says:

        how about a 5th category for the wrong knowns, i.e. things we think we know, but are wrong about? There’s a lot of that in Science…

    2. eub says:

      And one way to characterize engineering is as a knowledge and practice of what you don’t need to know. The bridge engineer does not need to understand the depths of materials science, because we’ve learned a way to describe a steel briefly that lets the bridge engineer disregard the rest.

      Until a bridge is built in a new fatigue regime or a new type of corrosion environment: how any given steel would perform was an unknown, and now it’s a known unknown. It’s an empirical question whether important unknown unknowns will crop up rarely enough that we can do anything useful.

      In trying to engineer on top of biology, these will crop up all the damn time.

  8. Anonymous says:

    If I had one of those promised autonomous vehicles (self driving cars), I would have more time to spend thinking about the bio problem on my commute … if I wasn’t overly worried thinking about a software or hardware glitch sending me into the opposite lane or over a precipice.

    1. ENES says:

      Funny you should bring up AV, check out the link and do a word search on drug or FDA. You may or may not agree with the views expressed but will be amused for sure….

      One of my pals (auto industry) moved to a new job, will be working on AV and wants to know more about the clinical development process; his point was safety testing of AV is not deterministic….go figure!

      Derek – I am long time reader of your blog and my regular lunch time reading; please accept my profound (belated) thanks.

  9. JBstein says:

    Do LEGO bricks communicate between each other ever ? do they funnel information and substances to form new bricks across the space ?
    arrogance does resemble at one stage to ignorance, no matter your CV

  10. Old Timer says:

    Derek, you should really update your LinkedIn profile. It would have been nice to know your new title as “Drug Industry Observer” without having to learn it from Scientific American.

  11. Pieinthesky says:

    View this article not as a “scientific” article, but as a prospectus/credential burnishing act for Pande’s role as a VC fund hawker. Think of all the millions that can be collected and funneled into vaporware bioAI and machine learning medicine start ups! Don’t you old-school drug development drones get it? Silly Valley has the vision and eyesight to discern cures that the rest of you are so hopelessly befuddled about!

    1. Doubter says:

      I tend to agree that this is a drum call for more investment to keep the AI/ML gravy train running. Orders of magnitude more data are needed as we are now hitting the wall everyone foresaw was coming, please comply and put the money on the table to keep the engine running.

    2. flintstone says:

      I think you’ve really nailed it–VCs are mostly about overhyping garbage “science” to drive up the valuation of the latest crap they are hawking. I’ve looked at a lot of VC-funded biotechs focussed on (anti)-aging, and they are simply ridiculous. So every now and then you’ll get a VC guy bashing real science and real outcomes while promoting “disruptors” and “mavericks” who are just eager but woefully ignorant science wannabees.

  12. I am always amused at these things, not because of the eloquently stated inapplicability to biology, but because of the author’s naivete about engineering itself.

    There’s some underlying assumption that we can reliably build software to spec, on schedule, simply be applying well-understood processes. If this were true, why would the software industry be inventing new processes almost as fast as it invents new software?

    In reality, every substantial software shop in the world is a mare’s nest of egos and stupidity. Software is invariably late, broken, and incomplete. Large swathes of code are invariably written based largely on guesswork followed by random tinkering until it kind of works.

    The idea that software development is some sort of swiss watch of repeatable precision is simply laughable.

    1. Dr. Manhattan says:

      You mean like this latest to hit Windows 10:
      “After consumers reported a number of problems with the latest major update to Windows 10 — including deletion of user files — Microsoft has pulled the update entirely to investigate.

      “We have paused the rollout of the Windows 10 October 2018 Update (version 1809) for all users as we investigate isolated reports of users missing some files after updating,” reads a note on the Windows 10 Update history page.”

      Imagine what would happen of a CRISPR system “engineered” to delete a particular deleterious gene in humans had a “bug”. Or worse, poorly understood cell processes forced it to behave in the wrong direction. Got a “software patch” for that?

      1. Isidore says:

        This reminds me of the following joke ca. 1990s: If automobile development had progressed along similar lines as the development of the personal computer, cars would cost $1,000, would get 200 miles to the gallon and at random intervals they would blow up killing all passengers.

      2. Anonymous says:

        This issue is of very serious concern to DARPA (think of ex-USA actors using these as offensive agents). How to detect, protect against, and remediate such “buggy” CRISPR biological weapons?

    2. Earl Boebert says:

      You beat me to it. I spent a large part of my career in the engineering of high-consequence systems (I post under my own name, you can look me up anywhere) and Pande fails to demonstrate an iota of understanding of what is required or how it is practiced. What we get is the standard SV hustle: “You’re doing it wrong, we’re rich and that proves we’re doing it right, do it our way.”

      If I may be so bold, I would suggest that the time of Derek and this community would be better spent on this paper:

      It raises interesting and important issues, is scholarly, serious, and informed by the author having grown up in the Soviet Union. By contrast, deconstructing Pande is as rewarding as deconstructing a late night infomercial.

      1. anon says:

        A bit long-winded at times, but a very good read. Thanks for sharing.

  13. ANonyMouse says:

    It’s yet another way the Grove fallacy is false. The massive advances in computer chips (Moore’s Law), storage (“Kryder’s law”), and genomics—all exponential decreases in cost, 1,000 times over a decade—come merely from 30% improvement year over year.
    Vijay Pande can’t seem to do simple math. 1,000x (or 1024x) in 10 years is a doubling every year. Moore’s Law was a doubling every 18 months. He’s (much!) smarter than that, so why the apparent fiddling with the numbers?

    The identification of biomarkers (chemical substances we can measure and then target) for disease is currently driven by discovery via a bespoke, one-off process—so the discovery of PSA for prostate cancer, for instance, does not suggest a biomarker for ovarian cancer. Introducing machine learning into the process, however, can turn this handcrafting into assembly-line production. … Entrepreneurs applying deep learning to medicine can use AI/ML and labeled data from generic Apple watch pulse data streams to accurately and precisely identify atrial fibrillation.
    Reasonable examples. Computers make it easy to identify bicycles in images, etc. This isn’t design, however. And knowing a 20-parameter composite metric that diagnoses atrial fibrillation doesn’t generally inform one on how to treat AF. Correlation doesn’t imply causation, and thus useful intervention.

    Most significantly, to my mind, Pande deftly dances around how “engineering”—the creating of reproducible, reusable parts, and implementation of design–build–test cycles—will address the key reason it costs ~$2 billion on average for each new drug.

    As Derek points out again and again, we don’t understand what are the biological parts gifted to us by billions of years of evolution, nor how they truly work together. We know only bits and pieces. Critically, we don’t really know which of those parts, if any, one can safely ignore during bottom-up design. (Top-down selective “design,” using evolutionary selection, is not, I think, what Pande is reaching for here.)

    Pande avoids writing what I just did. Succinctly: It’s the Biology we think we understand but really don’t that leads our billion-dollar biological “bridges” to keep falling down.

  14. anon says:

    ” How, for example, does anyone think that the Structural Genomics Consortium has been running through so many automated protein crystallography and X-ray structure experiments over the years? How did PCR get optimized to the workhorse procedure it is now? How did monoclonal antibodies move into industrial production? How, indeed, did something like Sanger sequencing get developed back in the early days? By breaking the problems down into pieces and optimizing them separately.

    Sounds easy. but in the end it is the same old stoty..

  15. z says:

    (Not sure if it’s because of spam moderation or captcha issues, but I tried to comment, and it didn’t seem to go through, so I’ll try again and see what happens….)

    This leads me to wonder, how do VC firms actually operate? I know very little about the field, so I’m probably wrong, but I imagine that the people working at these firms, the ones who are actually pushing investments like this, operate a lot like some (not all) of the people I’ve seen in drug discovery management–pursuing a kind of group-think, follow-the-leader approach where everyone jumps on the trendiest thing, and because everyone is doing it, this drives the short term value way up, meaning “cha-ching, bonuses all around!” And by the time it’s revealed that the emperor has no clothes, things have been re-organized enough that no one really remembers how they got there in the first place, and, hey, it was just the company’s money after all, no skin off my back, so on to the next trendy thing!

    1. Passerby says:

      Let’s be very clear: the romantic image of VCs actually caring about science or technology is just that – an illusion created by VCs to woo naive entrepreneurs. They’re really in it for the money, and the best analogy is with gamblers; they’re basically bookies betting on a horse. They could care less about the breeding or the biology of horses.

    2. MTK says:


      I wouldn’t be that cynical. Yes, all they really care about is return on investment, but in general I’ve been very impressed by VC’s, at least the ones that invest in drug discovery. They’re generally very experienced and knowledgeable in the field. It’s the VC’s that have asked the most probing and challenging scientific questions. They’re stewards of other people’s money and take their tasks seriously. Remember that they get paid in large part by how well the investments do, so they have to make good decisions.

      Now does that mean they don’t jump on trends? No, of course not. They’re not immune to human nature, but they don’t do that haphazardly and certainly not with a devil may care attitude if it doesn’t come to fruition. The successful firms do a lot of introspective analysis to figure out what things worked and where mistakes were made in order to hone their investment philosophies and decision making.

      That’s one of the lessons from the whole Theranos thing. All of the big life science VC firms took a pass on them.

      1. z says:

        Yes, I guess I was being too cynical. I shouldn’t use such a broad brush to characterize a range of different approaches. It’s just that we see so many people pitching ideas that are either unrealistically and meaninglessly vague and optimistic, or in a lot of cases that just sound obvoiusly scienticifically ridiculous, and yet it must be working for them. They’re getting money. The money must be coming either from people who don’t have the background and shouldn’t be investing in the sector, or from people who see these companies for what they are and still see some path for their own enrichment. It’s hard enough to be successful when dilligently focusing on good science, so it’s galling to see the success of hucksters.

  16. Old Pump Kicker says:

    The quote “they can now mass-produce many tests in a predictable, precise and repeatable manner” made me think of Theranos. I concur that he needs to cite examples of actual mass-produced tests. (Or clarify his definition of “now”.)

  17. dipthroat says:

    I would say that Pande’s argument has the same fallacy of the:”if cars would have improved like computers they would cost a fraction and be thousands times more capable”. The fallacy is that they are not equivalent problems to solve. They are different realities with different rules and constraints.

  18. Derek, I appreciate your challenge here of Pande’s over-the-top reductionist take on biology. (You have to give it to Pande for writing down what most of SV is thinking. On the other hand, he’ll now have to live with writing the line, “evolution is the ultimate algorithm!”) How true is your main contention that we just don’t know yet all of the components, even if it reinforces Pande’s modular metaphor. We only recently discovered circular RNAs. The comparisons to computers or radios or bridges–to any machines humans built is–as you say, looking back on 3.5 billion years of history through the modern industrial machine maker’s lens.

    You didn’t say anything about context. What about the reaction of the cell to surrounding cells, or of the organism to the environment? If you took all the particles that make up a nematode worm and put them in space, would we have a nematode worm?

    Would it be too esoteric to go on with the following point? When you write about biology, you are writing about science, not nature. Double stranded RNAs, micro RNAs, circular RNAs–these are abstract concepts, models. Biologists, let alone VCs, don’t appreciate the amount of metaphysics that they are doing. It takes a long time for empirical evidence to catch up.

  19. MikeC says:

    I’ll make a “look after your own house” type suggestion: Never mind creating a nematode, how about telling us how to repair a modern CPU? It shouldn’t be that hard, since the details of their construction and design are well understood. So it should be no problem for solid state engineers to explain how to diagnose and repair corrupted traces or gates in a given faulty CPU. Having to repair the actual elements as opposed to just bypass them, having to do so while the CPU is still in the computer, and completing the task without the OS being paused or rebooted won’t be a problem, will it? We won’t hear any arguments about “well, we didn’t design this to be fixed”, will we?

  20. Bin says:

    I think Pande’s most arguments hold true for bacteria, especially in E. coli. I would recommend people to read about the Cello paper ( That’s why he talked about the Lego thing.

    In E. coli, we indeed have reproducible, reusable parts, and implementation of design–build–test cycles. This is several years’ hard work in one of the top labs. Besides, genomic mining has shown its great power in discovering new parts and components, so I’m optimistic that the part category could keep expanding.

    Also it is a little misleading when we compare biological engineering with the development of hardware/software. Biology is very different from any other engineering discipline for following reasons:

    1. Currently we don’t have very good theory / model/ paradigm in biological engineering. So it’s hard to build more complicated biological systems based on other people’s work, like constructing better cars or CPUs with more transistor count. For software engineering, you can simply use some external library or add your code to the package as a developer. In biology, each lab has their own ‘toolbox’ and in most cases, only the people from the same lab can develop and use it.

    2. In biological engineering, we’re dealing with something that has been evolved for billions of years. That means that we must first ‘deconvolute’ the system and then engineer it. Physics principles haven’t evolved — we just believe Newton’s laws and it works perfect. It’s like nature has written code for several billions of years to build a operating system, with a lot of junk code and you’re a junior student want to understand it. You can not have another Newton’s law that works as a magic. Maybe someday we can ‘program’ in the same operating system (we already did) but I’ll not take it for granted that we could ‘write’ a similar operating system in the near future.

    1. loupgarous says:

      CRISPR is another “Lego” thing which is being advertised in E-books on Twitter.

      Can “The Complete Idiot’s Guide to Bioengineering be far behind?

      1. Bin says:

        Yes, exactly. And there’re more CRISPR variants being discovered by data mining.The next question, I think, is how we can use them in a systematic way.

  21. John Methot says:

    Sean Parker seems to have gotten the message. A STAT article about a Washington Post interview yesterday (linked via my name, above) contains this:

    ‘Parker said, “tech people coming from tech to biology so dramatically underestimate the complexity of the human body. It’s not designed by us. It doesn’t work in ways that make sense.”’

  22. eub says:

    Silicon Valley tech thinking can do certain things within biology, if it starts from “what do we really understand like Legos” and stay strictly to that. Isolated systems like “let’s do computation with RNA”, perhaps.

    But they have to be purely technique-driven, resisting the all temptation to veer even a little out of their lane to attack a problem because it matters, or would sell. Handling your RNA in vivo, for example, is way, way out of the lane. Anything with actual organisms in it, but especially anything with eukaryotes.

    Because yeah, have we found the Legos in that C. elegans yet?

    I’m reminded of a Lewis Thomas piece, that was talking about computer-controlled ICBMs rather than Silicon Valley VC investment, but both are talking about, how good does our formalism have to be to hope to understand a single cell? His modest proposal was a crash ten-year research program to make the computer understand that cellulose-digesting ciliate in termite intestines (or whatever its taxonomy is nowadays), in its structure and relations. His real meaning perhaps was to contrast life’s symbiotic mesh with megadeath calculations, but concretely it’s true: an engineering that can understand this organism would be one hell of an engineering, unlike anything we know.

  23. loupgarous says:

    I’ve seen how fallout and prompt weapons effects maps in simulations of large nuclear weapon exchanges work. Alex Wellerstein’s “NUKEMAP” neatly combines Internet maps with overlays of affected areas for blast, heat, prompt nuclear radiation and fallout.

    All the big megadeath calculations do is repeat what NUKEMAP does at the national and world scale. By the magic of computers, contourse of the area affected by well-understood nuclear weapons effects are applied to numerous nuclear detonations by lots of small calculations – all applicaitions of simple math.

    Nothing, not even TTAPS climatology calcuations laid over fallout and direct weapon effects calculations, are anything like so complex as the topography and function of a single cell. Cells are part of what C.S. Lewis once called “those vast and perilous estates, pulsating with the energy that made the worlds”.

    The map needed to comprehend the single cell is not yet complete. Even 20 years after genomics came into its own, entire organisms’ DNA sequences listed and recorded, we’re still finding out not just what DNA does in the cell, but what proteins and enzymes do in ways not mapped by DNA (the “proteosome”). Ten years for that turned out to be an optimistic estimate.

  24. Li says:

    Derek wrote “bolts are bolts”. Are they really? Even at HomeDepot, I can buy a couple of different types of plastic bolts, and a couple of types of metal ones. Brass, SS, Chromed, etc. And that’s ignoring heat treatment, alloy, etc. (HD has very low standards).
    Anyway, I’m not convinced that Computer Science will not provide some useful paradigms for biological systems. (I don’t hold a strong opinion either way.) Seems to me that a biological entity can be compared (somewhat) to the internet or maybe even an OS. (given the enormous number of different device configurations the OS has to handle). That said, we seem to continue to view a biological system as an entity, an object, rather than a dynamic (decaying) process (multi-process system, if system is interpreted loosely). The potential interactions for a single microRNA are similar (actually, much more simple, I think) than the potential interactions a web-surfer has. (Of course, adverse (system) interactions of the latter are limited) But it seems to me that it’s more likely that biology will have (has?) paradigms useful for distributed systems rather than the other way around for at least the near future.

  25. drsnowboard says:

    Perhaps mRNA being described as the ‘software’ of biological systems is a gentle nod to Silicon Valley style investors, given Moderna’s massive planned IPO and the equally huge returns to their executive team…

  26. David Edwards says:

    Even though my knowledge of biology is a long way behind that of some of the other contributors here, I still know enough to recognise that biology is not going to yield to hyperbolic assertions about “engineering”, of the sort being peddled by Pande, in the immediate future, and not only for the reasons Derek cites.

    Part of the problem is that software developers themselves frequently don’t realise what their software is capable of, until some evil genius hacker comes along and puts an exploit to malicious use. Software developers are being caught out in this manner on a pretty regular basis, despite the fact that they have, allegedly, a complete description of the system they’re working with, and a battery of testing tools to weed out unwanted side effects and bugs before the software goes “live”.

    The trouble with modern software development being, of course, that there are now systems in existence that are already too intricate to be understood in toto by a single human being. Which, worse still, were not developed by a single human being, but a sizeable aggregation of human beings, some of whom may not have documented everything they did at the level of detail required, or communicated to other members of the aggregation what they were doing in a manner enabling an exploit to be eliminated. The classic example of exploitable intricacies being PHP or JavaScript frameworks, which require those using them to rely upon the developers thereof to keep those frameworks robust and exploit-free. If those developers fail in this duty, then users of those frameworks are themselves working on a time bomb waiting to go off, even if those users exercise optimal diligence.

    So the idea that software development methodologies are going to be a magic wand curing the ills of biology, is not merely hopelessly optimistic, but potentially dangerous. At least if your software goes rogue on you, then the only casualties are your computers and the data they contain (such examples as flight management computers in airliners constitute a separate case for the purpose of my post). If “engineered biology” goes rogue, this could result in a pandemic wiping out billions of us. Not much chance of a patch downloaded from the Internet cleaning up after that.

  27. science is great
    I want to learn everything in this life

Comments are closed.