Slashdot Mirror
Hacking Biology
taatacgactc writes: "DARPA (DoD) is now supporting the development of a SPICE equivalent for biological circuits (aka BioSPICE). Best of all, it's to be "open source". Given the hyperexponential improvement in biological technology and the bioengineering efforts getting off the ground, fun stuff should be happening. Of course, there is the "dark side"." More information here. The submitter may be overstating the bit about "open source": the proposal says "All software developed as a part of the program will be open, in the sense that program performers and other DARPA authorized users will have the right to view, use, modify, and distribute code within the program authorized community. All derived works including revision, enhancement, modification, translation, abridgement and expansion of code will also remain open in this sense. ... The DARPA Director reserves the right to approve and exercise licensing arrangements depending on the context and the relevance to national security."
Only with models of what DNA actually does could we (maybe) figure this out. Right now the theory is mostly experimental -- knowledge that when people have done one thing, something else has occured. That's very crude.
An historical analogy might be with semiconductors. Semiconductors are not very complicated -- you add some cyanide to silicon, I believe, and you have a semiconductor that can act as a transistor. But no one thought to do this until their was a model (with quantum mechanics) that would imply that such a combination would have useful behavior. There's no quantum-level manipulation involved in making semiconductors, only quantum-level understanding.
In the same way, we may be able to do nearly everything we'd want to just by manipulating DNA. But we have to understand more than DNA to accomplish what we want.
Hmmm... to explain the nonlinear part: say we have an original DNA sequence of AA (this is entirely abstract). Once you find out what BA and AB do, you may still have no idea what BB does. Understanding what each isolated change does may not give you any understanding of what changes do in combination. To figure out what A and B mean is in many ways simply to figure out how to predict what they will do in combination. And DNA has so many possible combinations...
For that matter, is "dynamical" even a real word? (Check the title of the page on the "BioSPICE" link above - "Arkin Laboratory for Dynamical Genomics"...)
Here's a real word for these people - "Neologism."...
---
"They have strategic air commands, nuclear submarines, and John Wayne. We have this"
Hacker Public Radio is our Friend
I don't know about that, necessarily, depends how you look at it:
grep fluorescent_green_protein_gene JellyFish >> bunnygenes; cat bunnygenes > bunny_egg_cell; cat bunny_egg_cell | mama_bunny > glowing_baby_bunny.
Like the "core" Unix tools, our existing biotech tools ARE very "low-level", as you point out. We're certainly quite some way from the biotech equivalent of "higher-level" mix-and-match projects like php-gtk, but we've got some very useful building blocks to work with...including the project that is the subject of this article, which appears to be intended for building simulations of larger biochemical pathways...the next stage of biotechnology which we can't yet do much of, as you say. Sadly, this means I'm still probably at least a decade away from taking over the world with an army of Atomic Mutant Zombie Clones®
Incidentally, I'd be willing to bet that a determined and/or skilled individual can do a LOT with a bronze axe...but I think one level of analogy is plenty, so I'll avoid belaboring the point. Besides, I understand what you mean: Hype bad. Results good. :-) .
---
"They have strategic air commands, nuclear submarines, and John Wayne. We have this"
Hacker Public Radio is our Friend
Anything as complicated as DNA is non-linear -- small changes in source can result in large changes in result, and we cannot necessarily deduce from a fixed number of known source->result mappings what all combinations of source will do.[...]Only with models of what DNA actually does could we (maybe) figure this out.
I'm not quite sure what you're getting at here. If you mean we don't have more than a blurry understanding of chemistry involved in the proteins that are encoded in the DNA, you're absolutely right. It almost sounds as though you're implying that we don't know what DNA "does" at all, though, which is obviously not true. We know EXACTLY what expressed DNA "does" - it codes for proteins. We also know exactly what the translation of DNA-base-pair-triplets to amino acids is. Noncoding sections of DNA are obviously trickier, but some of them are reasonably well understood now as well (e.g. promotor sequences, binding sites for repressor proteins, etc.).
To figure out what A and B mean is in many ways simply to figure out how to predict what they will do in combination. And DNA has so many possible combinations...
I can hear the original poster hopping up and down with agitation at the DNA-centric message here even as we type :-)
The original poster's central [and correct!] point was that from a "functional" point of view, protein chemistry is more important than raw DNA, which I THINK is what you're getting at (if "A" and "B" in your example represent sequences coding for two separate proteins [or parts of one protein complex, like hemoglobin])
To clarify my analogy, I'm thinking of DNA as the "at, tr, grep, and cat" (ATGC, get it? Sorry, couldn't help myself. Watson and Crick made me do it...) of biology, compared to protein chemistry's "Xfree86, Mozilla, Sendmail, and Koffice". My point was just that while the tiny utilities are just a meager bit of the more broadly useful and complex applications, they are still extremely useful themselves, just as the DNA data available is itself of great (but very focussed) usefulness, even before we have better knowledge of the protein chemistry later. (we don't understand the source code well enough to sit down and rewrite XFree86 from scratch, but have a pretty good understanding of many of the little subroutines and how they might be put together into new "proglets"...)
---
"They have strategic air commands, nuclear submarines, and John Wayne. We have this"
Hacker Public Radio is our Friend
Definitely a good point, but I wouldn't underestimate the usefulness of simply being able to sequence better.
DNA Sequences alone can be/are useful for a lot of different research. Archaeologists and Anthropologists can use DNA sequences to help determine, for example, the relationship of one population to another and/or help trace migration of a population. Medical types can not only check for genetic diseases, but could also use DNA sequences to quickly check the identity of a pathogenic organism, if the sequencing technology becomes readily available enough. (Running a few PCR cycles is still much faster than trying to isolate and culture a pathogen from a swab). Zoologists and Paleontologists can use DNA sequences for similar purposes to what I mentioned for anthropologists and archaeologists, and can be handy for environmental research.
Until we understand the *function* of the proteins that are derived from these genes, all biotechnology can do is recombine the already existing technologies.Again, this is true...but don't forget that "recombining already existing technologies" can be pretty powerful and useful all by itself. (Heck, "recombining existing technologies" is, basically, a fundamental design principle of Unix-based systems, isn't it? I know MY Unix-based systems are extremely useful... I love my "|" key...)
I'm confident we'll be getting plenty of use out of DNA while we work on the harder problems of protein functions and chemistry
P.S. Thanks for reminding me...I've been meaning to download the folding@home client and throw some of my meager computing resources at helping out...
---
"They have strategic air commands, nuclear submarines, and John Wayne. We have this"
Hacker Public Radio is our Friend
I don't think it has been proven that knowing the
genome or pronome "code" allows to to manipulate
life to a serious degree.
Sure there have been results in some genetic
diseases, franken-food, and bio-identity.
But serious bio-hacking may require more than
just information sequences.
Jaron Lanier's recent "Half a manifesto" suggests
that information may not explain everything
and other aspects of reality may be acting there.
John Searle, the Berkeley philosopher has a similar
complaint toeards those who try to digitally emulate the mind.
Thinking that "information manipulation" explains
and controls everything may limit our understanding
of phenomena and ability to control it.
Don't get locked into this box.
Any development, any invention, anything in existance can be used to destroy.
And please don't start on about creating something like a human life, because I'll throw my first born at you.
For a good time call www.sawkie.com
Remember, DARPA are the folks who came up with DARPANET, which begat the Internet, which begat the Web, which begat /. and many other wonderful things. Mindless paranoia just because the military is involved would mean turning our back on a whole lot of very cool things. And when it comes to biotech, it's not just cool; it may very well be the thing that saves your life somewhere down the line. (In a related vein, e.g., anyone who's been in a serious accident any time in the lastt few decades has the military to thank for their care, because the modern EMS system is the direct descendant of the military medevac system, which was first developed in its modern form in Korea and was perfected in Vietnam.) Yeah, it's perfectly reasonable about the government doing Eeevil Things with the knowledge -- but on the whole I'm a lot _more_ worried about cancer and diabetes and heart disease and AIDS and all the other killer diseases which may be cured with the knowledge this project creates.
Also, as a former military medic (and current biotech worker) I can tell you that the military is often _directly_ interested in doing good things for people, not just seeing spin-off benefits (e.g., the way the personal computer can trace its lineage directly back to NASA and the USAF.) The military population -- active duty, retirees, and family members -- constitutes a large, diverse patient population with health problems ranging from arthritis to (obviously) bullet wounds. DARPA, USUHS, and other military research institutions have a strong and legitimate interest in medical advances -- which will first help soldiers, and then their families, but which will inevitably propagate to civilian medicine as well.
The correlation between ignorance of statistics and using "correlation is not causation" as an argument is close to 1.
"hyperexponential"? Michael, I think you are talking hyperbollocks.
Rich
If anyone has seen the Metal Gear: Solid Snake for Playstation, it revolved around this plot. Even had the DARPA Chief in it. Very interesting how fantasy and games now preceed real life by only a few years and months.
Apparently many people are not familiar with electronics.
SPICE is a program used in electrical engineering to simulate electrical circuits. As with any other simulation, it can be used to evaluate correctness, find logical and physical failures, test performance, find bottlenecks, verify timing properties, and other things.
Come to think of it, I guess you are trolling too, so I shouldn't be answering...but I can't resist. ;-)
Anyway, if you think that being willing to use force to defend innocent life is barbaric, I pity you. Human beings have a natural (or God-given, if that is how you look at it) right to life. The right to life implies a right to defend that life. The right to defend life implies the right to possess effective tools for defense. Firearms are very effective tools for personal defense.
BTW, police wearing bullet-proof vests is very cool. I have four cops in the family. I want them to have the best chance possible to stay alive. I also don't mind them having fast cars, as long as they use good judgement when deciding to continue a high-speed chase. (I don't own an SUV, but am considering buying one, if only to piss off the Sierra Club!)
Finally, with regard to your dim view of us yanks: Bite me.
"Rub her feet." -- L.L.
Couldn't any human being claim prior art if this were to be patented?
SPICE is a circuit simulator used by the EE types to design circuits.
As someone who is reasonably well conversant with this field (I've read many of the biological modelling papers, and heard Adam Arkin speak twice), I thought I would try and explain what's going on here. The DARPA call for proposals seems to contain two items - one on DNA computing and one on BioSPICE. As far as I can tell, they are two separate things. BioSPICE is an attempt to make a piece of software to model biological systems. It has everything you would expect: synthesis of RNA from DNA, translation of RNA into protein, chemical reactions, interactions, and so on. The goal is to use this software to model biological systems. The benefit of this is two-fold: one, you can test an existing biological system to see if the model agrees with the data; two you can use it to design new systems. However, the technology is very primitive. The only system that's beem modelled with any confidence is a piece of the phage lambda decision making circuitry (phage lambda is a virus that infects E. coli). There has also been some modelling of the cell cycle (the timing of DNA replication and cell division, but that was done more phenomenologically). As far as building novel biological circuits, this is still in its infancy: the only two things that have been built (both in the last two years) have been a flip-flop and an oscillator. So biology has a long way to go before it even gets to where electronic circuits were in the 40's.
The submitter may be overstating the bit about "open source": the proposal says "All software developed as a part of the program will be open, in the sense that program performers and other DARPA authorized users will have the right to view, use, modify, and distribute code within the program authorized community
Whew. The last thing we want is a bunch of B10-5kr1pt k1dd135 running around h4x0ring the neibourhood pets...
/bluesninja
As this article states 'Authorized' and 'licensed' too many times for me!
I can see it now, Microsoft has just now 'OpenSourced' Windows 2000. But only if you're stamped, approved, authorized, certified and licensed?
LFS. Have you built your system today?
It saves so much time in calculating the voltages and currents in circuits. Also, you can make perfect circuit diagrams. The Windows demo of DesignLab (which uses SPICE) is here; it's limited to 50 components, but that's enough to make simple circuits. I don't know if there's a Linux version of the GUI though.
"Ancillary does not mean you get to rule the world." --U.S. Circuit Judge Harry Edwards, speaking to the FCC's lawyer
just for your information
Rating the advances in biotechnology simply by looking at DNA sequencing improvements is not very smart..
We will not advance in understanding biology by simply looking at genes. That is like deciphering german ENIGMA codes, only to find out that you don't speak german. At the moment the advances are made in protein function (very difficult, and relatively slow progression).
Until we understand the *function* of the proteins that are derived from these genes, all biotechnology can do is recombine the already existing technologies.
If an experiment works, something has gone wrong.
Just wait till you get sued by the RIAA et all for having a program in your bodily fluids that extracts and decrypts DVDs ....
****
Court: You are hereby ordered to cease and desist any and all Semen production. Failure to do so will result ...