Supercomputer Performs Simulation of Virus 230
moller writes to tell us Red Herring is reporting that researchers from the University of California at Irvine and the University of Illinois at Urbana-Champaign have announced that they created a computer simulation of a virus. From the article: "Using one of the world's fastest computers at the U.S. National Center for Supercomputing Applications, located at the University of Illinois at Urbana-Champaign, the researchers ran a computer program devised to reverse engineer the dynamics of all atoms making up the virus particle and a tiny drop of water containing it." Nature also has an interesting write up on the research surrounding this project.
This is why we do scientific computing (Score:5, Interesting)
This is just fascinating, and precisely why we do high performance scientific computing. This quote piqued my interest in particular:
The model also shows that the virus coat collapses without its genetic material. This suggests that, when reproducing, the virus builds its coat around the genetic material rather than inserting the genetic material into a complete coat. "We saw something that is truly revolutionary," Schulten says.
So, by doing this simulation of a tiny span of time, the team was able to get new insight into the process of viral replication that would be extremely difficult to come by with experimental techniques. It also is fascinating, since we often think of viruses as little static particles that float around until they interact with a cell, and yet the simulation showed the surface pulsing. Very cool! -- Paul
Re:I dunno... (Score:3, Interesting)
There are millions to hundreds of millions of atoms in a typical virus. Here is interesting virus simulation info [indiana.edu]
Oh yeah, I know (hope?) you're joking, but modelling millions of atomic interactions is, as they say, nontrivial.
possible? (Score:5, Interesting)
And *That* is what computers are for! (Score:5, Interesting)
Unfortunately, we're still a few generations of supercomputer off from being able to simulate ribosomes (at which point most of the cellular machinery will be suitable for in-silicio biochemical investigation), but this is an excellent step along the way. It's also a good to showcase Schulten's group's work on efficient parallelization of complex simulations. He's had to solve a lot of algorithmic issues in order to be able to run that simulation, so this is not just an example of "wait for a bigger computer". If you check out their web-page http://www.ks.uiuc.edu/ [uiuc.edu], you'll find discussions of the underlying technology, which has required collaboration between biophysicists and computer science. My hat is off to them, especially as they not only achieved the proof of concept (we can simulate a small virus), but also gained biochemical insights (we didn't know they collapsed without the genetic payload). Bully for the Biophysicists!
Note: I don't work for them, but I admire the scale of simulations they do, and their willingness to make available to the community the tools they use.
shows how far off we are from simulating life (Score:3, Interesting)
No, this is scientific showboating. (Score:5, Interesting)
The researchers were using a technique called molecular dynamics, which attempts to model the movements of atoms in a 3D structure by integrating over Newton's equations. Force, however, is calculated using a coarse, empirical function of atom positions and their chemical properties. This model is weak, and it fails to produce physically-reasonable results on a whole variety of smaller problems, so it's an exaggeration to suggest that this simulation produced anything of physical or experimental relevance. And drawing strong physical conlusions from it? That's just crazy.
Before I get flamed by the MD crowd, I'll say that I am NOT suggesting that MD is useless. It's just that, it has a very short track record on problems of this size, and even in much smaller systems (i.e. fewer atoms), it's success rate is questionable. We can't even predict the dynamics of a single protein with this stuff -- it's absurd to suggest that it will work on an entire virus.
In short: don't be fooled. This experiment got into Nature because of its hubris and glamour, not necessarily because of its science.
P.S. I work in this field, so I'm posting anonymously.
You can run this yourself (theoretically) (Score:5, Interesting)
The modeling software they used is called NAMD [uiuc.edu], free open source "parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems" that will run on commodity clusters of tens of Linux PCs on gigabit ethernet. In other words, you too can run the virus simulation on your own Beowulf cluster, if you don't mind it taking some years to run. According to NCSA's own press release [uiuc.edu] about the virus simulation, it "only" took 35 processor-years, so if you have a 100 fast Linux PCs on a gigabit network lying around you can do it yourself in not much more than 4 months.
Re:No, this is scientific showboating. (Score:5, Interesting)
You give them too much credit... force field people compute *some* of the forces and ignore most of them. Long-range electrostatics are often omitted entirely (and people wonder why their RNA strands fall apart once it flops around a bit - I mean jebus, people!).
People do these simulations to "gain atomic level insight" into the problem; however, it's very rare to hear anyone say anything "insightful" about the chemistry AND do it in a believable way.
More "fascinating" simulations involve including a small region which is treated "quantum mechanically" and thus allow for bond formation/breaking; however, the QM models are so crude themselves that they need to be parameterized to get the "correct" answer. That's right - you heard it. It's the big secret. In order to get the more "trustable" simulations to produce something in the ballpark of remotely representing reality, you have to know "the answer" before you do the simulation and then teach the model to reproduce that "answer"... then you can write a paper and show that you're model "get's the answer" - and that's about the limit of "insight" that's often gained from these sort of simulations.
Living Things (Score:3, Interesting)
Re:No, this is scientific showboating. (Score:1, Interesting)
Re:Philosophical question (Score:3, Interesting)
Most large-scale patterns and processes in nature are stochastic and the outcome is one of a distribution of possibilities, similarly to quantum mechanics. In general, the same pattern can result from diverse processes (e.g., banded patterns of vegetation from dispersal dependence or topographical variation), while a single process may give rise to many different patterns (e.g., fire disturbances may result in even-aged forests or mixed-age forests).
Re:Philosophical question (Score:4, Interesting)
Philosophically, however, the question is interesting and has, as most interesting philosophical questions, been discussed before. Look up "Laplace's demon" sometime(the demon being the closest thing an 18th century mathematician could imagine to a supercomputer with access to infinite information): if one assumes that the world is inherently deterministic(from our point of view, this assumes the "hidden variables" interpretation of quantum mechanics) and proposes that some kind of being could in theory have access to unlimited information about it, then concepts such as time lose much of their meaning, since things are "destined" to happen before they actually do happen(and in what absolute sense can they then be said not to have happened already?).
Obviously, free will in the absolute sense is also non-existent from this point of view.
Re:No, this is scientific showboating. (Score:2, Interesting)
Uh, this is a troll. The AC has no idea WTF he's talking about. Either that, or he's an empiricist who can't stand to see computational science steal his thunder. Plenty of proteins have been modeled using QM/MM and had the results validated by empirical studies.
Re:No, this is scientific showboating. (Score:1, Interesting)
The cutoff is fine when there is a lot of water with its large dielectric; obviously it introduces errors (which I'm sure have been quantified by others) for other things but it's not the end of the method as we know it.
Re:No, this is scientific showboating. (Score:1, Interesting)
I second that this is scientific showboating... (Score:1, Interesting)
First off, this is not actually a virus perse, it's a satallite virion which can only reproduce during cocontaminant Tobacco-Mosaic Virus infection, and does not have the ability to infect anything itself. It is also only a fraction of the size of most viruses.
Second, Molecular Dynamics is a purely emperical, non-precise method for doing molecular simulation. Although to be fair, ab initio or Density functional theory approaches to this kind of a problem are impossible at this time. But there are some QM semi-emperical methods that are gaining popularity in the biology world.
Third, this is a tiny, minute time-scale for such a large chemical system. It's one thing to be working femtosecond timescale when you are simulating electron exchange between to small molecules, but on a system of this size means very little, especially when using an imprecise measure such as MM.
I guess I should also point out that it's not that hard to set up a simulation such as this when you are starting out with a hi-res x-ray crystal structure for you're starting coordinates. It's really just how much computational capacity you have.
Any thoughts?
-Ryan
Re:No, this is scientific showboating. (Score:1, Interesting)
(I'm not sure why you're bringing a stochastic simulation into this... if you're doing Monte Carlo, then the PME methods may not be necessary as the reciprocal space part of the Ewald sum can be formulated to scale well. If you're doing something like hybrid MC where you still have MD sequences, I presume you would still want to use PME.)
"Computing life" is possible (Score:3, Interesting)
This is an interesting coincidence because I used to reflect deeply on this exact subject a few years ago: what if a supercomputer could simulate a human ? I'll be honest here: I am literally _astounded_ to discover that this scientific team has successfuly simulated a virus. I didn't thought supercomputers were powerful enough for such a task. I just finished reading some articles about the experience and I now understand why this has been possible: they used some empirical functions instead of implementing exact physical laws (would have required much more computing power) and they also simulated the virus for only 50 billionths of a second. But still, they seem to have successfuly simulated life.
Most people don't realize the significance of this event, it means that given enough computing resources we could theoretically simulate humans ! One day we will have enough computing power to run such a simulation. And when it will be done, this human life simulation will have the potential to prove (or disprove) that humans are "just" a bunch of atoms following physical laws and nothing more.
This is huge. Think about it. I know this may sound sad, but personally I am convinced that any life form, including humans, is just that a complex assembly of atoms following physical laws, there is no soul, no afterlife, etc. This human life supercomputer will prove I am right :)