When Biology Becomes Software

All of life is based on the coordinated action of genetic parts (genes and their controlling sequences) found in the genomes (the complete DNA sequence) of organisms.

Genes and genomes are based on code -- just like the digital language of computers. But instead of zeros and ones, four DNA letters -- A, C, T, G -- encode all of life. (Life is messy, and there are actually all sorts of edge cases, but ignore that for now.) If you have the sequence that encodes an organism, in theory, you could recreate it. If you can write new working code, you can alter an existing organism or create a novel one.

If this sounds to you a lot like software coding, you're right. As synthetic biology looks more like computer technology, the risks of the latter become the risks of the former. Code is code, but because we're dealing with molecules -- and sometimes actual forms of life -- the risks can be much greater.

Imagine a biological engineer trying to increase the expression of a gene that maintains normal gene function in blood cells. Even though it's a relatively simple operation by today's standards, it'll almost certainly take multiple tries to get it right. Were this computer code, the only damage those failed tries would do is to crash the computer they're running on. With a biological system, the code could instead increase the likelihood of multiple types of leukemias and wipe out cells important to the patient's immune system.

We have known the mechanics of DNA for some 60-plus years. The field of modern biotechnology began in 1972 when Paul Berg joined one virus gene to another and produced the first "recombinant" virus. Synthetic biology arose in the early 2000s when biologists adopted the mindset of engineers; instead of moving single genes around, they designed complex genetic circuits.

In 2010, Craig Venter and his colleagues recreated the genome of a simple bacterium. More recently, researchers at the Medical Research Council Laboratory of Molecular Biology in Britain created a new, more streamlined version of E. coli. In both cases, the researchers created what could arguably be called new forms of life.

This is the new bioengineering, and it will only get more powerful. Today you can write DNA code in the same way a computer programmer writes computer code. Then you can use a DNA synthesizer or order DNA from a commercial vendor, and then use precision editing tools such as CRISPR to "run" it in an already existing organism, from a virus to a wheat plant to a person.

In the future, it may be possible to build an entire complex organism such as a dog or cat, or recreate an extinct mammoth (currently underway). Today, biotech companies are developing new gene therapies, and international consortia are addressing the feasibility and ethics of making changes to human genomes that could be passed down to succeeding generations.

Within the biological science community, urgent conversations are occurring about "cyberbiosecurity," an admittedly contested term that exists between biological and information systems where vulnerabilities in one can affect the other. These can include the security of DNA databanks, the fidelity of transmission of those data, and information hazards associated with specific DNA sequences that could encode novel pathogens for which no cures exist.

These risks have occupied not only learned bodies -- the National Academies of Sciences, Engineering, and Medicine published at least a half dozen reports on biosecurity risks and how to address them proactively -- but have made it to mainstream media: genome editing was a major plot element in Netflix's Season 3 of "Designated Survivor."

Our worries are more prosaic. As synthetic biology "programming" reaches the complexity of traditional computer programming, the risks of computer systems will transfer to biological systems. The difference is that biological systems have the potential to cause much greater, and far more lasting, damage than computer systems.

Programmers write software through trial and error. Because computer systems are so complex and there is no real theory of software, programmers repeatedly test the code they write until it works properly. This makes sense, because both the cost of getting it wrong and the ease of trying again is so low. There are even jokes about this: a programmer would diagnose a car crash by putting another car in the same situation and seeing if it happened again.

Even finished code still has problems. Again due to the complexity of modern software systems, "works properly" doesn't mean that it's perfectly correct. Modern software is full of bugs -- thousands of software flaws -- that occasionally affect performance or security. That's why any piece of software you use is regularly updated; the developers are still fixing bugs, even after the software is released.

Bioengineering will be largely the same: writing biological code will have these same reliability properties. Unfortunately, the software solution of making lots of mistakes and fixing them as you go doesn't work in biology.

In nature, a similar type of trial and error is handled by "the survival of the fittest" and occurs slowly over many generations. But human-generated code from scratch doesn't have that kind of correction mechanism. Inadvertent or intentional release of these newly coded "programs" may result in pathogens of expanded host range (just think swine flu) or organisms that wreck delicate ecological balances.

Unlike computer software, there's no way so far to "patch" biological systems once released to the wild, although researchers are trying to develop one. Nor are there ways to "patch" the humans (or animals or crops) susceptible to such agents. Stringent biocontainment helps, but no containment system provides zero risk.

Opportunities for mischief and malfeasance often occur when expertise is siloed, fields intersect only at the margins, and when the gathered knowledge of small, expert groups doesn't make its way into the larger body of practitioners who have important contributions to make.

Good starts have been made by biologists, security agencies, and governance experts. But these efforts have tended to be siloed, in either the biological and digital spheres of influence, classified and solely within the military, or exchanged only among a very small set of investigators.

What we need is more opportunities for integration between the two disciplines. We need to share information and experiences, classified and unclassified. We have tools among our digital and biological communities to identify and mitigate biological risks, and those to write and deploy secure computer systems.

Those opportunities will not occur without effort or financial support. Let's find those resources, public, private, philanthropic, or any combination. And then let's use those resources to set up some novel opportunities for digital geeks and bionerds -- as well as ethicists and policy makers -- to share experiences and concerns, and come up with creative, constructive solutions to these problems that are more than just patches.

These are overarching problems; let's not let siloed thinking or funding get in the way of breaking down barriers between communities. And let's not let technology of any kind get in the way of the public good.

This essay previously appeared on CNN.com.

EDITED TO ADD (9/23): Commentary.

Posted on September 13, 2019 at 11:40 AM • 31 Comments

Comments

KronosSeptember 13, 2019 12:52 PM

As a youth I often watched the old horror movies from the 1950's which featured giant creatures that became a menace after getting a dose of radiation. I can only wait for the real or imagined creatures/humans that result from biology programming gone awry.

Bob PaddockSeptember 13, 2019 1:42 PM

'Gene Bombs' that would only kill certain ethnic groups have been in discussion and hopefully not in implementation far longer than any of the technology referenced in the above.

For those that might like rabbit holes to alternate 'life':

In 1837, Andrew Crosse reported to the London electrical Society concerning the accidental spontaneous generation of life in the form of Acurus genus insects while he was conducting experiments on the formation of artificial crystals by means of prolonged exposure to weak electric current.


A. Crosse: The American Journal of Science & Arts; Vol. 35: 125-137 (January, 1839) ~ Reprint of Annals of Electricity, Magnetism, & Chemistry, vol. 2: 246-257 (January-June 1838)

Rex Research has more info on that.

Jesse ThompsonSeptember 13, 2019 2:28 PM

Cheers to the re-unification (physics term) of software, DNA, and the mind.

The latter already explored by Ghost in the Shell, Dollhouse, etc.

Following closely on the footsteps of the reunification of telephone, camera, computer, navigational system, music player, watch, etc.

The iPhone 13 will have apps to both cure Huntington's and instantly download kung-fu training into your corpus of personal memories.

Or just brainwash and cripple you. Whichever is of lower entropy I suppose.

vas pupSeptember 13, 2019 2:49 PM

Genetically modified mosquitoes breed in Brazil
https://www.dw.com/en/genetically-modified-mosquitoes-breed-in-brazil/a-50414340

After a field experiment between 2013 and 2015, genetically modified mosquitoes are breeding in Brazil. According to the researchers' original plan, all released mosquitoes and their offspring should have died.

The Yale research team around Jeffrey Powell warns that the newly formed mosquito population could possibly be more robust than the mosquitoes were before.

The authors conclude: "These results demonstrate the importance of having in place a genetic monitoring program during releases of transgenic organisms to detect unanticipated consequences".

Biologists critical of genetic engineering go one step further with their criticism, among them the Brazilian biologist José Maria Gusman Ferraz: "The release of the mosquitos was carried out hastily without any points having been clarified," Ferraz told the newspaper Folha de S. Paulo.

The Munich-based research laboratory Testbiotech, which is critical of genetic engineering, accuses Oxitec of having started the field trial without sufficient studies: "Oxitec's trials have led to a largely uncontrollable situation," CEO Christoph Then told the German Press Agency, dpa. "This incident must have consequences for the further employment of genetic engineering", he demanded.

JuhaniSeptember 13, 2019 3:43 PM

Are you sure biology should learn from it security?

From biology to IT I know only one case, where DNA was used to find a shortest path in graph. It took about one week and by direct computational energy use was cheap.

Biology is vastly different. It is like 1970-s IT security, very few have actual technology and it is like very expensive, some people have programmable pocket calculators. Not the least because those biology specialists have a very good education and charge a lot, my niece is one, she is engineering quite an interesting light activated molecule that releases the payload at for example cancer cell. That is postdoc level research using worlds top university research lab. It was like way more complicated than computer science, the biology has already some usable modules / science, but they are far from downloading a package / toolkit. Listening to niece, all the biohackers sound like babbling children, she was really hacking biology, that requires years of very hard work.
I know a little about it security and got a long current biology manipulation lecture from my niece and I fail to see any resemblance. That technology is not yet controllable, we lack the vocabulary to name problems.

DavidSeptember 13, 2019 3:56 PM

"All of life is based on the coordinated action of genetic parts (genes and their controlling sequences) found in the genomes (the complete DNA sequence) of organisms."

I'm sure that is wrong -- far too narrow a view of "life".

Given how complicated DNA is, I'm sure there was there were things we would have described as "life" before DNA was evolved.

And, even in modern times, there are interesting arguments as to whether or not viruses are "life" (though I'm inclined to think yes), and there are definitely viruses that do not have DNA.

Bruce SchneierSeptember 13, 2019 4:04 PM

@ David:

"'All of life is based on the coordinated action of genetic parts (genes and their controlling sequences) found in the genomes (the complete DNA sequence) of organisms.'

"I'm sure that is wrong -- far too narrow a view of 'life.'"

Maybe. I relied on my biologist coauthor for that sort of thing.

It's certainly close enough to being right for the purposes of this essay.

Alyer Babtu September 13, 2019 6:22 PM

Modern bioscience seems to be light years away from repairs for impediments that produce disease. If it can’t grasp the dynamics of the “standard” system, how can it proceed without catastrophe to modify it ?

JoeSeptember 13, 2019 7:35 PM

We know far too little about DNA and how it affects an organism to be sure what effects will occur. Our level of ignorance is profound in that we are constantly discovering that the "genetic code" is not a deterministic program. A recent article in Quanta Magazine described how genetically identical E. Coli show individual differences that are significant in terms of survival.
quantamagazine dot org/bacterial-clones-show-surprising-individuality-20190904/

This is worse than having unexpected outcomes because of mistakes in the code. It means that the exact same code can have different outcomes, which means testing it rigorously is much more difficult, if not impossible.

Schneier AdmirerSeptember 13, 2019 8:58 PM

Bruces essay seems justifiable on the face of it. But for my liking, it's too close to pouring petrol on the fire rather than water. It's shining light, attention and notoriety on a field, giving it credence in polite company, rather than shutting it down, making it illegal and ostracising all those who practice it. I'm not big on moral judgements but of that canon of things we would agree as evil, tampering with the very fabric of nature and Nature is entirely evil. And this site isn't the place for discussing ethics. It's just, well, weird finding this essay here.

Notably, Nicholas Nassim Taleb of Black Swan fame, has overwhelmingly strong feelings against biological engineering vs the continuation of the human race and planet earth.

Take all of Silicon Valleys 'save the world' hubris, and all of it's billionaire driven money wasting start up hubris, and the entire history of ComSec and InfoSec security failures and all the reasons for them, throw in Goldman Sachs and Deutsche bank executives, and US foreign policy, and times all of that by 100. And you may be getting just a little bit close to a vauge idea of possibly sniffing, what a future of biotech may be like.


Bruces final sentence:

'And let's not let technology of any kind get in the way of the public good.'

With the greatest of respect Bruce this is probably the strongest example of banality I recall encountering.

Public good? Technology 'getting in the way'? This in the context of
discussing a technology arguably more dangerous, more uncontrollable, and more mysterious than anything ever dreamed up before in history

Clive RobinsonSeptember 13, 2019 9:34 PM

@ Bruce,

Unlike computer software, there's no way so far to "patch" biological systems once released to the wild, although researchers are trying to develop one.

Err yes there is a way to "patch biological systems", the fact that we have not yet managed to bend it to our will is possibly just as well.

It is known that viral infections can cause cancer, and we are also begining to think other diseases like type II diabetes.

The fact that we have only so far identified "detrimental" virus "patching" does not preclude the possibility there is "benificial" patching.

Also whilst not patching all a persons DNA you might want to chat to your co-author about "Genetic Chimera that very occasionaly occurs in nature and also results from stem cell transplants. Thus you can have two or more different DNA signitures depending on what part of your body the DNA sample is taken from...

This has obvious security implications if you have had a bone marrow transplant. The blood you leave at a crime scene will not match the DNA taken from a cheek swab etc.

Thus as our bodies get more or less replaced at the atomic level every seven years a question arises in that if you replace some of your stem cells from a different person at what percentage do you cease to be the person you were?

It's kind of the "Grandfather's Hammer Conundrum".

    A century ago Grandpa purchased a hammer, over the years he, Pa and me have had to occasionaly replace the head and handle as they have worn out... But it's still Grandpa's hammer.

The more we progress down the road of Stem Cell Transplants, Genetic medicine and the like the more important this question becomes.

Gunter KönigsmannSeptember 14, 2019 1:11 AM

We have methods that in 99% of the cases modify the right genes. We already are producing wormable programs that should be able to extinguish only the species that are modified. Yesterday the news told that one of these experiments seemingly has escaped from the laboratory: The modified genes have starting to show in the wild. We know that there is vertical gene transfer between species, anyway, but don't know what mechanisms cause it.
And the code we edit is historically grown, undocumented, contains gotos, sections that are both executed and read as data, sections that are never read, but still magically change the outcome somehow and the publically available dev tools only check for known flu viruses. What could possibly go wrong?
Neanderthal sections that only one in a million persons contains and that might make harmless tools go haywire?
Pseudo-random sequences that look like a target? A highly-compressed (they all are) self-reproducing tool that somehow toggled a bit?

Wesley ParishSeptember 14, 2019 3:14 AM

One book that might be helpful in this discussion is Armand Marie Leroi's Mutants. It covers naturally occuring mutations and their effects. He covers matters like the sequence of events coding the limbs, for example, and what happens when things get out of sequence, or some steps are missed, and the like.

One thing that shouldn't be forgotten in discussing genetic engineering is that survival in the wild is far from assured. Any given mutation would have to be beneficial at least some of the time for an organization to survive. One of the most horrifying aspects of the anthropocene is that being able to survive in close quarters with homo sapiens/insipiens gives a survival advantage. And that includes bacteria.

Bill the Galactic Hero, RetdSeptember 14, 2019 1:37 PM

As I can personally attest, and as can my acquaintance Dr. Strangelove, adding foreign or altered parts, or better, pseudo-parts, even at the genetic level, to a natural organism does not create a new nature, but rather a monster, where the original nature is not augmented but impeded with something that takes ad hoc advantage of accidental possibilities inherent in matter. The impediment may turn out to have utility, in which case the monster is enslaved.

That is, one needs to maintain the distinctions between matter, form, substance and accident. This applies to artificial constructs such as machines, e.g. the hammer mentioned above is still grandfather’s hammer formally, in spite of material changes; as well as natures, where the biology renews all the matter of the body repeatedly over time. The genetics is not the nature and genetic alteration can not alter the nature. It is unlikely in the extreme that a new nature could be created.

See Aristotle’s definition of nature as "a source or cause of being moved and of being at rest in that to which it belongs primarily".

Eric JohnsonSeptember 14, 2019 2:26 PM

There are at least three other aspects of DNA "programming" that make it exceptionally hard for humans to create correct "programs". All the evidence I've seen throughout my career suggests that humans can be half-way decent at procedural programming. As in "First, do X, then Y, then Z, and if condition M, then perform Z again." That's not the only kind of programming, though. When it comes to DNA, the programming done there happens to be the kinds of programming that humans are worst at.

What I mean is that humans are generally good at single-threaded procedural programming, but not so good at parallel computing, event-driven design, concurrent operations, declarative programming, and emergent properties of systems. When it comes to parallel, humans usually get it wrong, because thinking about what happens when things happen at the same time gets complicated. Also, humans struggle with "event-driven" programming, as we tend to forget how events can misfire, and when things go wrong, how to respond with more events. Declarative programming is tricky, because it is then possible to come up with rules that depend on other rules, which then depend on the results of the original rules, and there's no clear answer. And finally, we still study simple "games" like "Go", and the "game of life", because they still elude easy explanations due to their emergent properties.

DNA combines all of these modes of programming. DNA drives molecule production in massively parallel operations throughout an organism, on the basis of responding to events. And yes, it is declarative, in the sense that "make X when Y" seems to be the operating principle. And all of these modes of programming combine to make life an emergent property.

Which makes me think that we're going to suck at programming DNA for a very long time.

RichardSeptember 14, 2019 9:38 PM

"Programmers write software through trial and error. Because computer systems are so complex and there is no real theory of software, programmers repeatedly test the code they write until it works properly."

There is no real theory of software? What's computer science?

WinterSeptember 15, 2019 7:41 AM

"Unlike computer software, there's no way so far to "patch" biological systems once released to the wild, although researchers are trying to develop one."

It is called "Crispr Cas"

https://en.wikipedia.org/wiki/CRISPR

For the rest, whatever we have "Designed" in biology up to now is laughable in terms of danger by what the living world has developed itself.

The most dangerous creature in the world, besides humans, is the mosquito (the genera Aedes, Anopheles, and Culex) that kill around 750,000 people each year and infects ~700 million each year with some horrible disease. We still have not found a way to combat them.

Yes, we can hack DNA, but we are very far from improving its deadliness. Biological weapons labs have today mostly killed those who worked there and lived in the neighborhood.

Clive RobinsonSeptember 15, 2019 1:21 PM

@ Winter,

Re mosquitoes and their range of diseases and,

We still have not found a way to combat them.

There are a number of reasons for that,

1, They are not --or were not-- a Western problem (yet).

2, The drug trials and approvals will be expensive (and problematic since certain drug developers were caught doing trials in countries where the costs are way way less).

3, As there are rules about drugs for pandemics (which these deseases are) which alow the effected countries to manufacture regardless of patents, there would be no cost recovery.

Which means the ongoing research certain drugs companies are known to be doing because some of these desieses are now becoming visable in the south of America, is going to be kept on the shelf untill a US --rigged health care-- market opens up and prices can be set at a hundred USD a day or higher (for a tablet or injection that will cost maybe a dollar or two tops to manufacture).

If you think that is unlikely hsve a look at what happened to the price of epi-pens for people with life threatening alergies oh and you might want to read this,

https://www.newyorker.com/magazine/2017/10/30/the-family-that-built-an-empire-of-pain

WinterSeptember 15, 2019 2:16 PM

"There are a number of reasons for that,"

This is valid for individual diseases. But for every disease we "Conquer" two new ones will rise, transmitted by mosquitos. E.g., zika did not exist a few years ago.

However, getting rid of, i.e., making extinct, of any of these genera of mosquitos is quite a different matter. That is like exterminating all rats. Won't happen soon.

If ever.

ThinkSeptember 15, 2019 2:26 PM

@Winter

Hopefully one day CRISPER will be used to help identify genetic sequences or codons associated with cancer cells and then insert the proper instruction(s) in only those cells with those unique genetic sequences to program the infected cells gracefully to die (apoptosis) - the chore then will be to clear the necrotic tissue and let the patient recover.

The technology is certainly more mature in places with less moral controls.

https://en.wikipedia.org/wiki/Dolly_(sheep)

Think China

https://www.technologyreview.com/s/613890/a-third-crispr-baby-may-have-already-been-born-in-china/

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6331330/

https://www.nature.com/articles/d41586-019-01580-1

Could Crisper be used against the human race in a biological hack - most certainly - instructions used in working with Crisper held on insecure internet connected computers at research locations and universities could be altered by state sponsored attacks and unwitting researchers in one country who think the are building a benign genetic experiment to instead build a deadly pathogen.

WinterSeptember 15, 2019 2:33 PM

"Could Crisper be used against the human race in a biological hack - most certainly"

Finally, the Zombie virus becomes reality!

Nah, more likely, someone will engineer a deadly virus, like a fast working AIDS or a slowed down Ebola virus. Will maybe kill billions, including its creators. But all humans? Not likely.

Alyer Babtu September 15, 2019 2:43 PM

@Think et al

identify genetic sequences or codons associated with cancer cells and then insert the proper instruction

It’s not clear this approach would work. The body is set up to manage for health and correct itself, so the question is why did the management fail ? Editing might just push the problem down the road.

Also what about epigenetics, where different behavior is seen even when the DNA is he same ? And what is the programming analogy ?

AbeSeptember 16, 2019 2:00 AM

Regarding "computer science": the problem is human beings trying to handle enormous complexity. The complexity of a non-trivial computer system is beyond the IQ of any team of sufficient size to handle the required workload. There are all sorts of approaches towards handling the complexity but they are very ivory-tower and based on implicitly flawed models of flawed human beings. So you end up with the MacOS beachball or various Windows issues. We are only so bright but our pride blinds us to this -- we are far too self-satisfied with our very recent technical achievements as a race. Look up the Ariane 5 first launch disaster for a canonical example of what results in complex computer systems from very simple mistakes.

David LevineSeptember 16, 2019 10:59 AM

Imagine attacking the software at a near-future factory that creates DNA-based therapies. Just as a virus takes over the (re)production process within the cell, a computer virus can take over the production process inside of a bioengineering factory.
• Easy: Make the drugs ineffective
• Harder: Make the drugs harmful to the person who takes it
• Hardest: Make the “drugs” a container for a pathogen, so a drug company is distributing a biological weapon you designed.

TheoSeptember 16, 2019 4:05 PM

Science is a method of studying something you don't understand. Computers are every bit as valid a subject for science as dark energy.

RealFakeNewsSeptember 18, 2019 5:37 PM

then use precision editing tools such as CRISPR to "run" it in an already existing organism

CRISPR has been exposed as one of the least reliable methods for gene editing. Other far superior systems were swept aside after a "publicity bomb" was dropped a few years ago screaming how great it was, drowning out the better methods.

CRISPR relies on a process that is malleable (in biological terms, the process itself can modify the process). It results in gene sequences that appear to work, then unexplained changes occur and things go awry quickly. It has a very high failure rate, and in addition, what can only be described as instability (seemingly random mutations occur generations later as a result of the process).

SamSeptember 19, 2019 1:48 PM

@Winter

Zika has been known since the 1950s, it is not new.

Zika is the perfect example of the most dangerous human invention: efficient global transportation. Air travel and containerized shipping, and globalization in general, amplify a small regional disease a global problem.

This has been repeated over and over with all kinds of disease and invasive species, like West Nile, Zebra Mussel, even feral dogs and cats.

What is needed is a large tax on long-distance travel and imports to account for the biological impacts. Of course, academics and politicians are the type to jet off to a foreign country for a 2-day conference and spend their frequent flyer miles for a weekend in Paris, so nothing will ever happen.

Leave a comment

Allowed HTML: <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre>

Sidebar photo of Bruce Schneier by Joe MacInnis.