lowest energy question

Message boards : Number crunching : lowest energy question

To post messages, you must log in.

AuthorMessage
Scott14o

Send message
Joined: 7 Apr 06
Posts: 24
Credit: 2,147,598
RAC: 0
Message 22310 - Posted: 12 Aug 2006, 1:47:47 UTC
Last modified: 12 Aug 2006, 1:48:29 UTC

If a user predicts the lowest energy, what does that do? I've read Dr. Baker's stuff, but it just says that we're trying to find the lowest energy, not what it is used for (at least what I saw).

Also, what if the lowest prediction really isn't the LOWEST level that there is?

oh yeah and when will we get the results from CASP?
ID: 22310 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Feet1st
Avatar

Send message
Joined: 30 Dec 05
Posts: 1755
Credit: 4,690,520
RAC: 0
Message 22311 - Posted: 12 Aug 2006, 1:57:20 UTC - in response to Message 22310.  

If a user predicts the lowest energy, what does that do? I've read Dr. Baker's stuff, but it just says that we're trying to find the lowest energy, not what it is used for (at least what I saw).

Also, what if the lowest prediction really isn't the LOWEST level that there is?

oh yeah and when will we get the results from CASP?


Scott, welcome to Rosetta!

Not sure on CASP, I'd say "Thanksgiving timeframe" from what I've read elsewhere.

The lowest energy prediction is most probably the one that is closest to the correct (i.e. "native") structure of the protein. So, even if it is not "THE lowest" one might find if they searched more, it's the best prediction you've got... and it's often close ENOUGH to correct to be useful to understand how it will interact with other proteins and whether or not it might make a cure for some specific disease.

If you picture a complex twisty turny rollercoaster, if you correctly predict where to get on, that's about all you need to know. And so if your prediction of that portion of the protein is correct, and that is the portion that will be interacting with other proteins in diseased cells, that's the important part.
Add this signature to your EMail:
Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might!
https://boinc.bakerlab.org/rosetta/
ID: 22311 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Keith Akins

Send message
Joined: 22 Oct 05
Posts: 176
Credit: 71,779
RAC: 0
Message 22313 - Posted: 12 Aug 2006, 2:32:17 UTC

Hello Scott.

Another aspect of this whole excercise is to gradually reduce the number of models per protein to get at the lowest energy and closest model to the native structure. As more is learned about the energy landscape of each protein, The team can adjust and tweak the methods and protocols used in prediction. They work on ways to reduce invalid energy domains so that gradually the models returned hopefully will begin to cluster around the true native structure.

The better your prediction of a known structure is, the better your prediction of an unknown designer protein is likely to be.

But Like Feet1st said, only the active portion of the protein is really important.
ID: 22313 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile SOAN
Avatar

Send message
Joined: 27 Sep 05
Posts: 252
Credit: 63,160
RAC: 0
Message 22372 - Posted: 13 Aug 2006, 0:05:54 UTC - in response to Message 22313.  


Another aspect of this whole excercise is to gradually reduce the number of models per protein to get at the lowest energy and closest model to the native structure. As more is learned about the energy landscape of each protein, The team can adjust and tweak the methods and protocols used in prediction. They work on ways to reduce invalid energy domains so that gradually the models returned hopefully will begin to cluster around the true native structure.


So let me see if I got this right. Does that mean...

An intitial set of predictions are made without any information about the shape of the protein. Then the Dr. Baker and friends look at the predictions with the lowest energies to get a good idea of the rough structure and can narrow the search further by knowing having that rough information.

Basically, are there stages of prediction for each protein or just separate, non-overlapping approaches?


ID: 22372 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Feet1st
Avatar

Send message
Joined: 30 Dec 05
Posts: 1755
Credit: 4,690,520
RAC: 0
Message 22373 - Posted: 13 Aug 2006, 0:40:07 UTC
Last modified: 13 Aug 2006, 0:41:41 UTC

I believe the stages of prediction you are picturing are actually occuring in the software on your PC. But yes, the game here is to take the sequence of Amino Acids, which is comparatively easy to determine chemically... and predict what shape the beast will take in nature. Since the proteins are beyond microscopic, you can't just pull one out and look at it. The structures are presently determined via either X-ray crystallography, using an X-ray setup that covers acres; or MRI using some of the strongest magnets in the world. Either way, the current methods are slow and expensive when you consider there are 100s of thousands of proteins we'd like to know about.

Dr. Baker was kind of teasing us when CASP first started. He said:
Here is its amino acid sequence:

MSFIEKMIGSLNDKREWKAMEARAKALPKEYHHAYKAIQKYMWTSGGPTDWQDTKRIFGG
ILDLFEEGAAEGKKVTDLTGEDVAAFCDELMKDTKTWMDKYRTKLNDSIGRD

can you tell from this what the three dimensional structure and function
of this protein are?


That is exactly the task he's taking on. Because of the complex atomic level interactions amongst atoms, many shapes are possible for this... but in nature, only one shape is ever found. So... what is it? That's what Baker lab is working to understand. And we're helping crunch the numbers for them to study how new approaches to the problem are helping them get better and better at predicting that shape.
Add this signature to your EMail:
Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might!
https://boinc.bakerlab.org/rosetta/
ID: 22373 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile dgnuff
Avatar

Send message
Joined: 1 Nov 05
Posts: 350
Credit: 24,773,605
RAC: 0
Message 22865 - Posted: 18 Aug 2006, 1:28:22 UTC - in response to Message 22373.  
Last modified: 18 Aug 2006, 1:29:06 UTC

Dr. Baker was kind of teasing us when CASP first started. He said:
Here is its amino acid sequence:

MSFIEKMIGSLNDKREWKAMEARAKALPKEYHHAYKAIQKYMWTSGGPTDWQDTKRIFGG
ILDLFEEGAAEGKKVTDLTGEDVAAFCDELMKDTKTWMDKYRTKLNDSIGRD

can you tell from this what the three dimensional structure and function
of this protein are?


That is exactly the task he's taking on. Because of the complex atomic level interactions amongst atoms, many shapes are possible for this... but in nature, only one shape is ever found. So... what is it? That's what Baker lab is working to understand. And we're helping crunch the numbers for them to study how new approaches to the problem are helping them get better and better at predicting that shape.


I posted something similar, and explanation of what we're doing in another thread.

I'm going to post my vision of the "Why?" in this thread.

Why do we care what shape a protein folds into?

Two reasons come readily to mind.

If we understand what shape a protein should fold into, it'll help with research into Alzheimers, which is caused when proteins don't fold correctly. Knowing why this happens may be pivotal in finding a cure.

Also, David Baker has emphasized that the algorithms used by Rosetta in the "Amino Acid sequence to shape" problem lend themselves to the reverse problem: Shape to amino acid sequence." Meaning it we want to make a protein with a particular shape, what sequence of amino acids do we want? Solving this allows the construction of custom proteins, which, among other things, could be leveraged to provide a vaccine for HIV.
ID: 22865 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Number crunching : lowest energy question



©2024 University of Washington
https://www.bakerlab.org