October 21, 2005 5:23 PM PDT

Lab computer simulates ribosome in motion

Using a computer to simulate the interaction of 2.6 million atoms, Los Alamos National Laboratory researchers have re-created a tiny slice of one of the most fundamental genetic processes of life.

The lab simulated how a cellular machine called a ribosome follows genetic instructions to construct a complex molecule called a protein out of building blocks called amino acids. With 768 processors of LANL's 8,192-processor ASCI Q machine running for about 260 days, the researchers created a movie of the process. Previous views had shown only static snapshots.

"Experiments have been able to come up with snapshots of the ribosome. We're trying to create a movie of what happens between those snapshots," said Kevin Sanbonmatsu, a molecular biologist and the project's principal investigator.

The movies could be significant for research into antibiotic medicines. Antibiotics work by gumming up the ribosomes, and a movie showing a ribosome's function could show a larger range of targets than static images, he said.

The task wasn't simple. Researchers had to model the physical interactions of each of 2.64 million atoms--about 250,000 in the ribosome itself, but most involving water molecules inside and outside it. The simulation resulted in a movie that is 20 million frames long, he said.

ribosome gallery

In reality, however, the ribosome behavior that they simulated takes only 2 nanoseconds, or 2 billionths of a second--too short to even be labeled as "fleeting."

Ribosomes are fundamental to life. They're "thought to be one of the oldest artifacts from the beginning of life that we can study today," Sanbonmatsu said. "If you compare the (genetic) sequence of ribosomes in humans and in bacteria, it's very, very similar. Most of the core of the ribosome is identical in every organism that's ever been sequenced."

The new research illuminates previously known biological mechanisms that begin with genetic information stored in DNA. That information is transferred into biological reality through a multistep process. Proteins--complex molecules such as hemoglobin to transfer oxygen in blood, or insulin to help metabolize sugars--are made of a chain of amino acids, and DNA encodes the order of the amino acids for each type of protein.

To create a protein, the dual strands of DNA are temporarily unzipped to permit the creation of a single-strand copy of the genetic information, called messenger RNA. The messenger RNA is then processed by the ribosome.

The ribosomes connect, in the appropriate amino acid, to the growing chain that forms each protein. Amino acids are carried into this molecular factory by tiny packages called transfer RNA.

What the LANL researchers think they've found is a corridor in the ribosome that screens out the inappropriate amino acids from the sea of transfer RNA.

"What we've discovered in the simulation is that this is a possible mechanism to accept or reject transfer RNA," Sanbonmatsu said. "This corridor acts like a gate."

Next, the researchers will try to experimentally verify the simulation's results, simulate antibiotic interactions with the ribosome and model how the ribosome moves step-by-step along the messenger RNA strand.

This simulation used six times as many atoms as the previously largest model known. That scale is significant, Sanbonmatsu noted.

"This allows us to look at more-realistic and physically relevant systems," he said. A ribosome itself is a "huge complex of messenger RNA and 50 proteins. Most things in cells are complexes of RNA and proteins."

Such simulations are increasing dramatically in sophistication. A machine called Blue Gene/L at LANL's sister lab, Lawrence Livermore National Laboratory, is scheduled for a ceremonial unveiling Thursday. Blue Gene/L is currently the world's fastest supercomputer, sustaining calculations at the rate of 136.8 trillion per second compared with 13.8 trillion for ASCI Q. Blue Gene/L performance is expected to roughly double this year as all its processors are installed.

12 comments

Join the conversation!
Add your comment
Biology and computers Intersect at last!
Thank GOD and the Bonmatsu team for this finding! This is one of the finest examples of the success of federal funding for fundamental research.

For 20-some years, I've been asked "how did you get into the computer business when you studied biology?" Well, here's the answer, AT LAST.

We now know that it was advances in software that made the mapping of the human genome possible in less than a lifetime. Now it's a computer simulation that's telling us how ribosomes work.

Isn't computer science WONDERFUL? Now if only people will stop asking me the OTHER question: how did you get into computers when you're not an engineer!
Posted by MariaBizDev (2 comments )
Reply Link Flag
I thought they already had
I thought the big industries that used lots of computing power were things like oil, theoretical physics and biotech research, so this story is basically an example of the third one in a matter of speaking in a roundabout way sort of?
Posted by steviesteveo (29 comments )
Link Flag
Biology and computers Intersect at last!
Thank GOD and the Bonmatsu team for this finding! This is one of the finest examples of the success of federal funding for fundamental research.

For 20-some years, I've been asked "how did you get into the computer business when you studied biology?" Well, here's the answer, AT LAST.

We now know that it was advances in software that made the mapping of the human genome possible in less than a lifetime. Now it's a computer simulation that's telling us how ribosomes work.

Isn't computer science WONDERFUL? Now if only people will stop asking me the OTHER question: how did you get into computers when you're not an engineer!
Posted by MariaBizDev (2 comments )
Reply Link Flag
I thought they already had
I thought the big industries that used lots of computing power were things like oil, theoretical physics and biotech research, so this story is basically an example of the third one in a matter of speaking in a roundabout way sort of?
Posted by steviesteveo (29 comments )
Link Flag
I don't get it
I don't get it. They used thousands of processors to make render a 3D sequence? What exactly they didn't knew by know?

Silly PR
Posted by BillyMcBean (7 comments )
Reply Link Flag
It isn't rendering a pre-determined sequence
In essence, all the modellers would have had to go on were the
co-ordinates of the atoms at fixed stages in the process (e.g. the
start, maybe the middle, and the end) and it would be necessary
to predict the sequence in between. The processor power is
needed to predict where each atom is likely to be at each point
in time in the entire process. Hence the massive computing
power needed as there are hundreds of thousands of atoms in a
ribosome and its co-factors in the process (tRNA, amino acids,
etc).
Posted by No invasion of privacy (52 comments )
Link Flag
I don't get it
I don't get it. They used thousands of processors to make render a 3D sequence? What exactly they didn't knew by know?

Silly PR
Posted by BillyMcBean (7 comments )
Reply Link Flag
It isn't rendering a pre-determined sequence
In essence, all the modellers would have had to go on were the
co-ordinates of the atoms at fixed stages in the process (e.g. the
start, maybe the middle, and the end) and it would be necessary
to predict the sequence in between. The processor power is
needed to predict where each atom is likely to be at each point
in time in the entire process. Hence the massive computing
power needed as there are hundreds of thousands of atoms in a
ribosome and its co-factors in the process (tRNA, amino acids,
etc).
Posted by No invasion of privacy (52 comments )
Link Flag
Fighting Mad Cow disease and simulating memory
This type of resolution is needed for fighting
Mad Cow disease and other 'protein' orientated
problems.
Also great for Science Fiction buffs like me who
are interested in simulation of organics. Stuff
like physically experiencing cyberspace.
Uploading yourself.
A college in Switzerland is using Blue Gene to
simulate the human brain to fight mental
disorders as well.
Posted by Blito (436 comments )
Reply Link Flag
Fighting Mad Cow disease and simulating memory
This type of resolution is needed for fighting
Mad Cow disease and other 'protein' orientated
problems.
Also great for Science Fiction buffs like me who
are interested in simulation of organics. Stuff
like physically experiencing cyberspace.
Uploading yourself.
A college in Switzerland is using Blue Gene to
simulate the human brain to fight mental
disorders as well.
Posted by Blito (436 comments )
Reply Link Flag
 

Join the conversation

Add your comment

The posting of advertisements, profanity, or personal attacks is prohibited. Click here to review our Terms of Use.

What's Hot

Discussions

Shared

RSS Feeds

Add headlines from CNET News to your homepage or feedreader.