1) Message boards : Number crunching : All WUs suddenly failing - output file absent (Message 37018)
Posted 48 days ago by Profile nate
FYI http://www.gpugrid.net/forum_thread.php?id=3770&nowrap=true#37017
2) Message boards : Number crunching : NOELIA WU (Message 37017)
Posted 48 days ago by Profile nate
Hello all.

The NOELIA_SH2eq workunits were all failing, so those have been cancelled.

There some other groups that follow the naming pattern NOELIA_TRPXXX. Those continue to run, because from what I can tell the ones that are being sent out now do not have a problem. The first post in this thread (http://www.gpugrid.net/forum_thread.php?id=3770&nowrap=true#36977) refers to a group that was cancelled, fixed and resent earlier in the week. However, if you continue to have problems with those please let me know here.

3) Message boards : Number crunching : CPU work units (Message 37007)
Posted 49 days ago by Profile nate
I believe there are some on the way right now. Stay tuned!
4) Message boards : Number crunching : NOELIA WU (Message 37006)
Posted 49 days ago by Profile nate
Hey guys,

We're looking into it. Unfortunately Noelia is away right now so we may have to cancel them for now. Thanks for catching it and pointing it out.

5) Message boards : News : WU: CMYBKIX (Message 36975)
Posted 54 days ago by Profile nate
Hi all,

I have submitted a bunch of new workunits to the short queue. This should fix the dry spell we have had there for the past week.

The workunits are named "NATHAN_CMYBKIX". In these simulations I am investigating the binding of a small disodered protein named cMyb to another protein named KIX. Disordered proteins and disordered domains (as opposed to nicely structured ones like hemoglobin) are very important in protein-protein interactions. Mutations in disordered proteins are very common in cancers, amyloid fibril diseases like alzheimer's and parkinson's, and probably a lot of other disease we don't know about. So, with these simulations we will be taking some first steps towards understanding some key things about these types of interactions. Lessons we learn from this system should help us work on many, many others.

6) Message boards : Server and website : No cpu tasks? (Message 35906)
Posted 121 days ago by Profile nate
Strange. We have a lot of workunits in the line-up, so there shouldn't be an issue. We'll look into it.
7) Message boards : News : WU: CMYB_equil (Message 35905)
Posted 121 days ago by Profile nate
Those were the corrected ones. There is currently a CMYB_equil3 batch as well, that also seems to be going without problems. These simulations are part of a much bigger batch that will come soon.
8) Message boards : News : WU: CMYB_equil (Message 35853)
Posted 123 days ago by Profile nate
And indeed there was a problem, so I have stopped them and I will resend them corrected. Sorry about that. For some reason checks before submission did not flag the problem. Very frustrating.
9) Message boards : News : WU: CMYB_equil (Message 35850)
Posted 123 days ago by Profile nate
Hello all,

I have submitted some simulations of a small peptide to the short queue. This is a special type of simulation, and it is also the first time I am running them, so there may be errors with them. Please let us know what you see. There should be pass through quickly, however, so hopefully they don't cause too many problems.

10) Message boards : News : Uncovering the role of membrane lipids in modulating enzyme activity (Message 35000)
Posted 163 days ago by Profile nate
Hello everyone,

At the end of last year, we updated the badge system to include several new publications we had just published. Today we want to highlight one of those pieces of work.

Starting in Mid-2011 (an eternity ago), we began collaborating with a research group in Italy to better understand something strange they had been seeing in their experiments. Specifically, they were seeing that when they modified the composition of lipids in a cell membrane, the activity of an enzyme named FAAH increased. They came to us to try to better understand why that was happening. Through simulations performed here on GPUGRID we discovered that, contrary to expectations, the lipids modified the function of the enzyme by directly interacting with it and its endogenous ligand. It had been expected that structural changes in the membrane would be the culprit, but that was not the case. Additionally, we report the first unbiased binding of a lipid to a membrane protein. This is one of the first experiments of this kind ever performed, and was possible only because of the compute power we have here at GPUGRID. You can read more about it on the science page here, as well as check out the video of the binding event.

Please note that (almost) all the papers we publish are available on our lab webpage. They can be found here. The versions of the documents there are the pre-print versions, basically meaning they are not formatted to look pretty. This is due to restrictions placed on us by the publishing companies. The text and figures are identical to those in the "pretty" version published in the journal, however. We are working on putting out copies of any papers missing from that page.

Lastly, we remind you that GPUGRID accepts donations beyond compute time, and if you're interested you can see how to do so on our Donations page. Of course, the crunching time donated by all of you is greatly appreciated.

Thanks and onward,

Next 10