Skip to Content

Google’s Calico Moves Into Reality

Google’s Calico venture, the company’s out-there move into anti-aging therapy, has made the news by signing a deal with AbbVie (the company most of us will probably go on thinking of as Abbott). That moves them into the real world for sure, from the perspective of the rest of the drug industry, so it’s worth taking another look at them. (It’s also worth noting that Craig Venter is moving into this area, too, with a company called Human Longevity. Maybe as the tech zillionaires age we’ll see a fair amount of this sort of thing).
On one level, I applaud Google’s move. There’s a lot of important work to be done in the general field of aging, and there are a lot of signs that human lifespan can be hacked, for want of a better word. The first thought some people have when they think of longer lifespan is that it could be an economic disaster. After all, a huge percentage of our healthcare money is already spent in the last years of life as it is – what if we make that period longer still? But it’s not just sheer lifespan – aging is the motor behind a lot of diseases, making them more like to crop up and more severe when they do. The dream (which may be an unattainable one) is for longer human lifespans, in good health, without the years of painful decline that so many people experience. Even if we can’t quite manage that, an improvement over the current state of things would be welcome. If people stay productive longer, and spend fewer resources on disabling conditions as they age, we can come out ahead on the deal rather than wondering how we could possibly afford it.
Google and AbbVie are both putting $250 million into starting a research site somewhere in the Bay area (and given the state of biotech out there, compared to a few years ago, it’ll be a welcome addition). If things go well, each of them have also signed up to contribute as much as $500 million more to the joint venture, but we’ll see if that ever materializes. What, though, are they going to be doing out there?
Details are still scarce, but FierceBiotechIT says that “a picture of an IT-enabled, omics-focused operation has emerged from media reports and early hiring at the startup”. That sounds pretty believable, given Google’s liking for (and ability to handle) huge piles of data. It also sounds like something that Larry Page and Sergey Brin would be into, given their past investments. But that still doesn’t tell us much: any serious work in this area could be described in that fashion. We’ll have to use up a bit more of our current lifespans before things get any clearer.
So I mentioned above that on one level I like this – what, you might be asking, is the other level on which I don’t? My worry is what I like to call the Andy Grove Fallacy. I applied that term to Grove’s “If we can improve microprocessors so much, what’s holding you biotech people back”? line of argument. It’s also a big part of the (in)famous “Can a Biologist Fix a Radio” article (PDF), which I find useful and infuriating in about equal proportions. The Andy Grove Fallacy is the confusion between man-made technology (like processor chips and radios) and biological systems. They’re both complex, multifunctional, miniaturized, and made up of thousands and thousands of components, true. But the differences are more important than the similarities.
For one thing, human-designed objects are one hell of a lot easier for humans to figure out. With human-designed tech, we were around for all the early stages, and got to watch as we made all of it gradually more and more complicated. We know it inside out, because we discovered it and developed it, every bit. Living cells, well, not so much. The whole system is plunked right down in front of us, so the only thing we can do is reverse-engineer, and we most definitely don’t have all the tools we need to do a good job of that. We don’t even know what some of those tools might be yet. Totally unexpected things keep turning up as we look closer, and not just details that we somehow missed – I’m talking about huge important regulatory systems (like all the microRNA pathways) that we never even realized existed. No one’s going to find anything like that in an Intel chip, of that we can be sure.
And that’s because of the other big difference between human technology and biochemistry: evolution. We talk about human designs “evolving”, but that’s a very loose usage of the word. Real biological evolution is another thing entirely. It’s not human, not at all, and it takes some time to get your head around that. Evolution doesn’t do things the way that we would. It has no regard for our sensibilities whatsoever. It’s a blind idiot tinkerer, with no shame and no sense of the bizarre, and it only asks two questions, over and over: “Did you live? Did you reproduce? Well, OK then.” Living systems are full of all kinds of weird, tangled, hacked-together stuff, layer upon layer of it, doing things that we don’t understand and can’t do ourselves. There is no manual, no spec sheet, no diagram – unless we write it.
So people coming in from the world of things that humans built are in for a shock when they find out how little is known about biology. That’s the shock that led to that Radio article, I think, and the sooner someone experiences it, the better. When Google’s Larry Page is quoted saying things like this, though, I wonder if it’s hit him yet:

One of the things I thought was amazing is that if you solve cancer, you’d add about three years to people’s average life expectancy. We think of solving cancer as this huge thing that’ll totally change the world. But when you really take a step back and look at it, yeah, there are many, many tragic cases of cancer, and it’s very, very sad, but in the aggregate, it’s not as big an advance as you might think.”

The problem is, cancer – unrestrained cellular growth – is intimately tied up with aging. Part of that is statistical. If you live long enough, you will surely come down with some form of cancer, whether it’s nasty enough to kill you or benign enough for you to die of something else. But another connection is deeper, because the sorts of processes that keep cells tied down so that they don’t take off and try to conquer the world are exactly the ones, in many cases, that we’re going to have to tinker with to extend our lifespans. There are a lot of tripwires out there, and many of them we don’t even know about yet. I’d certainly assume that Larry Page’s understanding of all this is deeper than gets conveyed in a magazine article, but he (and the other Google folks) will need to watch themselves as they go on. Hubris often gets rewarded in Silicon Valley – after all, it’s made by humans, marketed to humans, and is rewarded by human investors. But in the biomedical field, hubris can sometimes attract lightning bolts like you would not believe.

25 comments on “Google’s Calico Moves Into Reality”

  1. Twelve says:

    “A lot of tripwires” in human biology is putting it mildly – it’s essentially all tripwires, hidden mines, and bizarre hacks. There are a few apparently straightforward malfunctions with relatively easy fixes, but they only serve to seduce the unwary into believing they understand it all.
    If the Googlers, flush with cash and having intimations of mortality, want to send some $ our way, bully for them – more jobs in this field are always welcome.
    As for their chances of success, have a look at the Google driver-less car. It had been hyped as practically ready for the market, but it’s far from that. It must be told every minute detail of the route it is to travel, or it just retreats to the side of the road. A new stop sign would be ignored, unless it was painstakingly entered into the car’s database. Google’ colossal information warehouses are the world’s biggest hammer, and that’s their idea of how to solve every problem. Expect something similar from Calico.

  2. Anon says:

    I just can’t help but remember:
    To me this is analogous to a Nobel laureate in one field making strong claims in another (and years later everyone realizes this person was wrong). Being an expert in A doesn’t necessarily mean you will even be average at doing B.
    If they can’t use their data to mine their bread-and-butter for flu info…how are they going to mine the genome?
    Techie-Styled-Confidence. On in the Bay area.

  3. NoDrugsNoJobs says:

    A major problem we face in our industry is the iterative turn around time for hypothesis testing. Given the empirical nature and safety requirements of our work, it often takes a decade and usually more from the time we discover or focus on a pathway until we know it works. In contrast, mechanical/electric proptotypes can be tested in a very short time and multiple iterations explored rapidly. often times testing is not so much required as the output can be highly predictable and the design of the system is all that it takes to know the system will work. we’ve actually done pretty well given that in just several decades we’ve made fundamental inroads to the treatment of some diseases. In the long run I am sure we will solve many more of the challenging problems facing us but it will be a long run indeed. How to prove you are increasing lifespan without a controlled study to demonstrate that? Given a normal healthy lifespan is quite long, the clinical data of proof is likely to take multi-generations simply to test a singly hypothesis.
    Daunting but important work!!

  4. Lune says:

    I think it’s at least an unwarranted assumption and also probably unhelpful to characterize living systems that we don’t understand as objectively bizarre, weird, or tangled. The most elegant chip designs are bizarre, weird, and tangled *to me*, but from the correct perspective they are exactly as complex as they need to be.
    There have been multiple cases in biology where tangled and seemingly excess complexity ended up being just the thing that was needed, and I tend to think that the sooner we start thinking of the complexity of a living system as hiding some elegant solution to a hard problem, rather than as superfluous strangeness, the faster we’re likely to progress in understanding it.

  5. nonnie says:

    I’ve heard that sirtuins are good targets.

  6. Tomas says:

    There are examples of man-made systems that no one understands anymore. We use biological approaches to query them because they are opaque to a large extent. Google itself consists of thousands of services dependent on each other in complex ways, engineered by disparate teams.
    It can be a huge effort to deprecate (shut down) a service and make sure the downstream services continue to function.

  7. Anonymous says:

    Part of our problem in Drug Discovery is that there are so many inputs and so much data (and SO MANY things you have to get right to get a drug). Maybe Google (or IBM/Watson) can help us out with that data management/analysis bandwidth challenge. I’d be willing to bet they can come up with some cool data-mining approaches. I just hope the tools make it out to the public domain rather than being held back as a proprietary advantage for Calico.

  8. Boo says:

    Derek’s remark that “Maybe as the tech zillionaires age we’ll see a fair amount of [aging related research]” reminded me of something I’ve been thinking about in idle moments recently. I spent a good chunk of my career doing growth factor research aiming towards possible cancer or diabetes applications. Now that I’m heading towards the end of my career, I wish that I had researched ways to address the age-related decline in visual acuity. The thought never occurred to me until I was nearly 50. I wonder what other “advanced” researchers might wish they had pursued in their youth, had their youthful selves the perspective (if not the actual knowledge) that comes with age.

  9. johnnyboy says:

    When I searched around recently on what Calico was doing research-wise to approach the aging problem (apart from hiring big names as executives and making deals), all I found was that they were doing whole-genome sequencing on a bunch of old people. I’m sure that they are doing other things (well, at least I hope so), but this should give you an idea of how far they are from getting anything concrete into clinical trials.

  10. johnnyboy says:

    Oh and by the way, just wondering: how do you do a clinical trial to test an anti-aging drug ? Give the drug for 10 years ? And which endpoints would you evaluate ? They’d obviously have to be surrogate – gonna be fun to get the FDA to approve those.

  11. mike says:

    Just read the Labeznik BIologist/Radio paper (again). His fundamental logical flaw is that in his analogy, engineers are God. They invented the radio from nothing. They built it from nothing, and they know how every part works because they put them there–like God (if you believe in God) and cellular life. How would today’s engineer understand a Klingon cloaking device? A trans-warp transporter? A defective light saber? Probably the same way as a biologist trying to understand apoptosis.

  12. Anonymous says:

    “Evolution doesn’t do things the way that we would. It has no regard for our sensibilities whatsoever. It’s a blind idiot tinkerer, with no shame and no sense of the bizarre, and it only asks two questions, over and over: “Did you live? Did you reproduce? Well, OK then.””
    Shame we don’t develop drugs like this, instead of trying to understand and design everything like some egotistical god.

  13. Anonymous says:

    Speaking of man-made systems and complexity, does anyone understand the financial system? I mean really?

  14. milkshake says:

    @13: There is always a evolutionary biology explanations. The boom and bust cycle – see the Wikipedia entry on lemmings…

  15. aa3 says:

    The Larry Page quote about curing all cancer only adding 3 years to life expectancy is a brilliant point. I remember a few years back I read that even if all disease was cured, we would add 12 years to life expectancy. And frankly most of those 12 years would be in elderly people. At some point returns are diminishing in any paradigm of technology.
    It is at that point that the broader industry has to branch out into new frontiers. This is a big issue for the pricing power of the pharmaceutical corporations. The aging process is a process that affects 100% of the population. The market for any treatment that reversed some of the damage caused by the aging process in any one of our systems, would have a market of anyone who could afford it. It is why I believe sooner or later technologies to reverse damage from aging will become the largest industries in the world.
    Until the advance of those technologies themselves hit diminishing returns, (when almost all age related damage is being reversed, with minimal side effects at low cost).
    I do not know if the silicon valley engineering approach will make headway in biotech, but I wouldn’t rule it out. Ultimately it is about logic and creativity.. and the ability to coordinate large numbers of intelligent people to work together towards the target goal. And silicon valley has shown an uncanny ability to deliver advances.

  16. Paul D. says:

    ‘Evolution doesn’t do things the way that we would. It has no regard for our sensibilities whatsoever. It’s a blind idiot tinkerer, with no shame and no sense of the bizarre, and it only asks two questions, over and over: “Did you live? Did you reproduce? Well, OK then.”‘
    This reminds me of one of the recent clean-tech startups (another example of a Silicon Valley invasion of another tech culture). This startup, Siluria, decided to use high throughput screening techniques to make industrial catalysts. 70,000+ trials later they apparently have a practical, robust catalyst for the selective oxidative coupling of methane to C2 compounds at comparatively mild conditions.
    I’m wondering what other industrial chemical processes their approach could be applied to.

  17. Howard Furst says:

    A little nuance of “curing cancer will only add 3 years to average life expectancy” is that it’s averaged out over the whole population. For all of the individual people who get cancer fairly early in life, a cure would seem a bit more meaningful.

  18. Morten G says:

    @16 Paul D. makes a very valid point. Where Calico and research institutions like it could contribute a lot is by bridging the academic low-throughput science and the industrial high-throughput engineering. Especially with things like reproduction and finding the limits of where a new result is no longer valid.
    I’m not a computer science person or a mathematician but isn’t the complexity of a chip more polynomial where the complexity of a biological system is more exponential?

  19. NoDrugsNoJobs says:

    I think pharma has been exploiting high throughput processes for quite some time, I’d venture pahrma led materials exploration in this regard rather than the other way around

  20. simpl says:

    Discussed two off-the-scale but genuine projects over lunch with a clinical researcher, comparing the Google contact lens which monitors blood sugar with an anti-cancer antibody amplification project. Opinions differed but, importantly, you want the engineers to be impressed with the engineering, and the clinicians impressed with the medical results, not vice versa. (Chemists are onlookers here).

  21. Paul D. says:

    @19 Siluria’s approach came from Angela Belcher at MIT. She and her team have been using phage display libraries to make materials. Siluria grows their catalysts on engineered phages (M13 if I recall correctly), with the wide variety of surface proteins acting as templates for catalysts with different structures. The phages also cause the resulting catalysts to have a useful “nanowire” structure after the organic bits are burned away.
    Obviously this depends on biotechnology, but I don’t think it was directly pharma inspired.

  22. cancer_man says:

    “curing cancer will only add 3 years to average life expectancy”.
    Another nuance, a *big* nuance, is that this is based on the life expectancy around the year 2000. As life expectancy significantly increases, three years will become four , five , and six if a cure is not found for cancers.
    Cures would also save a great deal of money that could be funneled into places like Calico.

  23. sgcox says:

    The main problem,IMHO, with people trying to get into biology from hi-tech successful experience is the lack of understanding of the development. Mac or IPhone or DeepBlue are the same, atom by atom from the point of its construction. The human (or mice for that matter) are go from single cell to a fully formed and functional organism capable of very complex behaviour. Yes, it is all encoded as instructions in our genome but it is more like
    “All the human life’s a stage, and all the genes and siRNA merely players: they have their exits and their entrances; and one gene in his time plays many parts, his acts being seven ages…”
    Nothing remotely comparable we have in even the most complex computer systems.

  24. formerGNE says:

    You failed to mention that the CEO of Calico is not Sergey Brin (nor was Sergey the brainchild behind this), the CEO is Art Levinson, the founder and former CEO of Genentech. He has recruited, in my opinion, one of the smartest medical-scientists as the Chief Scientific Officer, Hal Barron, from Genentech as well. They have begun recruiting the top academic scholars from the field of BOTH aging and cancer (your article uses comments to suggest the company is just focused on aging, but it’s focused on research in both aging and cancer). Not mention its recently announced licensing agreement around P7C3 compounds.
    If it was just Google behind this venture, I would completely agree with your article. But it’s not … it’s some of the brightest minds in biopharma tackling some big diseases (CNS, cancer, aging) with a TON of money.
    We should be cheering them on!

  25. Michael says:

    Part of what underlies Page’s comments about cancer is actually likely to be *exactly* the fact that cancer – like atherosclerosis and Alzheimer’s – is intimately tied up with aging. Because all of these are diseases of aging (that is, of progressive accumulation of damage to the cellular and molecular structures over time), effective intervention in aging necessarily implies delaying the initiation and/or decelerating the course of age-related disease. As regards cancer, specifically: every intervention in aging that works in rodents today — Calorie restriction; dramatic restriction of the amino acid methionine; mutations in production of IGF-1 or of growth hormone or IGF-1 signaling; rapamycin and some of its analogs — concomitantly substantially delays tumorigenesis, reduces tumor burden and growth, and reduces death *from* and *with* cancer.

Comments are closed.