Welcome to the MacNN Forums.

Shaktai · May 25, 2002, 01:27 AM

05/24/2002
- The CASP5 structure prediction experiment is now underway, and will be proceeding throughout the course of the summer. Thus, starting towards the end of next week (Thursday approx.) we will prematurely terminate the present protein we are working on to begin phase III, on the first CASP target. We will resume phase Ib, with the next large protein of known structure, once the CASP 'prediction season' has passed. From the users' point of view, this simply means an auto-update will occur and RMSDs will no longer be computed (since the structure of the new protein is not known).

Okay let's get ready. Everyone make sure your autoupdate.cfg's are setup and ready to go. This is what everything has been leading to. From the noise in the forums, a lot of teams may be gearing up.

I know I have been missing in action here recently, but I will have everything I got back on-line for this. Will the Dfold process work to actually predict unknown proteins?? This is the big test. No more guessing if it is worth it or not. Time to bet all the marbles and throw everything we have into it. Not just for the team but more importantly for the science.

Get your game faces on... ... ... ...this is the big leagues now.

[ 05-25-2002: Message edited by: Shaktai ]

Grozni Majmun · May 25, 2002, 01:45 AM

Whoa!!!
Good to see you back in fighting form, Shaktai. We were getting worried here. My autoupdate is ready to go. "Unfortunately" I am also graduating shortly and probably moving to some remote corner of the world with failing electricity supply and internet scarcity, so I won't be able to contribute much. But I am sure Team MacNN will do fine. We've got some great people. In fact, we're: The single most amazing group of Mac users ever assembled for anything. Ever.

Happy crunching everyone.

[ 05-25-2002: Message edited by: Grozni Majmun ]

rkadowns · May 25, 2002, 10:53 AM

!

RIP

jbcool · May 25, 2002, 12:05 PM

Originally posted by Shaktai:


Okay let's get ready. Everyone make sure your autoupdate.cfg's are setup and ready to go.

I know I have been missing in action here recently, but I will have everything I got back on-line for this. .... Time to bet all the marbles .......

Get your game faces on... ... ... ...this is the big leagues now.

[ 05-25-2002: Message edited by: Shaktai ]

Thanks again Shaktai!! Just when I was having fun shaking things up over at the folding@home team you are going to make me start up the distributed folding clients so they get updated.

I lost my marbles a long time ago.

Can I use these bowling balls.

Hey don't get so mad that you turn blue in the face.

I'll hunt for my game face and try to have it back on this next week.
But these guys over at the folding@home team are starting to get serious.
Guess it is time to come back anyway with RIP driving that steam roller at me and slipjack almost out of sight.

The whole team has really picked thing up since I went away to play.
Great job everyone.

Shaktai · May 25, 2002, 03:22 PM

Originally posted by jbcool:
Thanks again Shaktai!! Just when I was having fun shaking things up over at the folding@home team you are going to make me start up the distributed folding clients so they get updated.

What? After all this time and you are still only #4?

I lost my marbles a long time ago.

Can I use these bowling balls.

Bowling balls are fine, just watch where you are tossing them.

I'll hunt for my game face and try to have it back on this next week. But these guys over at the folding@home team are starting to get serious. Guess it is time to come back anyway with RIP driving that steam roller at me and slipjack almost out of sight.

And here I was going to talk with slipjack and RIP about getting together and creating a combined member called slip, RIP and shake, that could do a cool 800,000 a day.

Actually, it is not slipjack or RIP you need to worry about. Super M is building up one heck of a head of steam, and looks to be determined to rule the roost. (Way to go Mikkyo) About the only person who is safe for now is Enola because she has such a big lead. And of course those of us who he has already steamrolled over.

enola · May 25, 2002, 04:33 PM

Originally posted by Shaktai:

Super M is building up one heck of a head of steam, and looks to be determined to rule the roost. (Way to go Mikkyo) About the only person who is safe for now is Enola because she has such a big lead.

You make it sound like I'm going away.

Good job Mikkyo!!! These folks need someone else to slap 'em around a bit and wake 'em up. You see, from where I'm standing my arms don't reach that far back! Heh heh!

On a more serious note: CASP 5!!! This is exciting. It's the real deal. I'm going to have to make sure my crew knows about the autoupdate.cfg file. So once the "prediction season" is over is that when results get published and we know if dFold works or not?

[ 05-25-2002: Message edited by: enola ]

Shaktai · May 25, 2002, 10:30 PM

<blockquote>:<hr />Originally posted by enola:
 So once the "prediction season" is over is that when results get published and we know if dFold works or not?<hr /></blockquote>The prediction season runs I think until the Fall. Now after that we have to wait until the actual protein make ups are determined by other more traditional (and expensive, time consuming) methods. What CASP does is test various predictive methods. The actual proteins are having their makeup determined now, so at this point, no one knows what they look like. All that is known at this point is their chain of amino acid strings. The object is to come as close as possible to to predicting the real thing. If the Distributed Folding team comes close on multiple proteins, or even "wins" by being the closest on 1 or more, it will help to validate their methods. If they do poorly, then it means they will have to reavaluate their scoring methods. If by some grace they should "sweep" Casp 5 in the categories they have chosen, it will mean we are onto something really really good. In December, the participants will meet at the Asilomar Conference grounds near Monteray, CA and the results and winners will be revealed.

If I understand correctly, after the Prediction season, we will return to the next stage of testing, but will have to wait to hear the results.

Makes me wish I could be in Asilomar again. Many, many years ago, I had the priviledge of speaking at some Philosophical Round Tables held there. It is a beautiful, and very peaceful place.

enola · May 29, 2002, 11:32 PM

From the distribfold project News section:

"- Just a reminder, an auto-update to change proteins for CASP will take place on Thursday, May 30, at around noon EST or later. The present protein will be terminated at this point and NOT continued, so for users with lots of computers, please upload any cached hordes of work NOW. Thank you."

Make sure you've got your autoupdate.cfg in place. Crossing fingers that the transition goes smoothly.

The_Equivocator · May 30, 2002, 01:06 PM

It's started! If you don't have your autoupdate.cfg file set, start updating those clients!

Unfortunately, the new protein appears to need a lot more computation that the old one. 182 residues have to be calculated per fold! <img border="0" title="" alt="[Eek!]" src="eek.gif" />

Shaktai · May 30, 2002, 01:47 PM

<blockquote>quote:<hr />Originally posted by The_Equivocator:
It's started! If you don't have your autoupdate.cfg file set, start updating those clients!

Unfortunately, the new protein appears to need a lot more computation that the old one. 182 residues have to be calculated per fold! <img border="0" title="" alt="[Eek!]" src="eek.gif" /> <hr /></blockquote>It's a big un!!!! <img border="0" alt="[Hmmm]" title="" src="graemlins/hmmm.gif" />

mikkyo · May 30, 2002, 02:10 PM

This new update doesn't contain altivec aware code does it?

Scotttheking · May 30, 2002, 02:23 PM

<blockquote>quote:<hr />Originally posted by mikkyo:
This new update doesn't contain altivec aware code does it?<hr /></blockquote>Nope

Wow this new one is a slow sucker.
We are going to need a lot more power to move up now.

wheeles · May 30, 2002, 02:38 PM

<blockquote>quote:<hr />Originally posted by Scotttheking:
Wow this new one is a slow sucker.
<hr /></blockquote>[lurk off]
182 steps for each unit rather than 76.

I had a couple of problems with the autoupdate failing. Might just be a dodgy modem connection, but it's worth doublechecking all your boxes updated properly. Once I had one updated I copied the changed files to my other crunchers. They are: sspred.val protein.trj foldtrajlite (and whatsnew.txt).
[lurk on]

rkadowns · May 30, 2002, 02:39 PM

<blockquote>quote:<hr />Originally posted by Scotttheking:
 <blockquote>quote:<hr />Originally posted by mikkyo:
This new update doesn't contain altivec aware code does it?<hr /></blockquote>Nope

Wow this new one is a slow sucker.
We are going to need a lot more power to move up now.<hr /></blockquote>It will slow everyone else too assuming the hit is linear. You know what they say about people who assume though...

Does anyone have any idea if this new process still fits within the L2 cache of the G4?

Shaktai · May 30, 2002, 05:03 PM

<blockquote>quote:<hr />Originally posted by rkadowns:
 <blockquote>quote:<hr />Originally posted by Scotttheking:
 <blockquote>quote:<hr />Originally posted by mikkyo:
This new update doesn't contain altivec aware code does it?<hr /></blockquote>Nope

Wow this new one is a slow sucker.
We are going to need a lot more power to move up now.<hr /></blockquote>It will slow everyone else too assuming the hit is linear. You know what they say about people who assume though...

Does anyone have any idea if this new process still fits within the L2 cache of the G4?<hr /></blockquote>Didn't get a chance to look at it much before work, but it was slow on my Athlon too. Probably by late tonight I should be able to give a comparision between my 4 machines (Athlon 1900+, Celeron 1.3 ghz, Power Mac G4 800 and iBook 600 G3) That is if I remembered to switch everything over???

enola · May 30, 2002, 09:55 PM

Well, the transition didn't go too badly. I've been in contact with most of my crew today and things went okay for them. The G4 QS at home updated fine but the iBook died.

I have the client running off of a RAM disk on the iBook and after much head scratching ( I kept getting these awful fatal core errors ) I figured out it was running out of space. So I created a bigger RAM disk and now it's cool.

Oh, and I'm getting a new computer at work. I'm going from a 500MHz P3 to a 1.7GHz P4. I'll be installing the client on that box as soon as I can.

Scotttheking · May 30, 2002, 11:09 PM

Well at least one updated, one didn't.
Can't reach the other two.
I have to turn off the setiQ and move it somewhere with a monitor, and the sis's imac is sleeping, and she's locked in her room right now.
At least I'm ordering a 1GHz Tbird tomorrow

enola · May 31, 2002, 12:18 AM

<blockquote>quote:<hr />Originally posted by Scotttheking:

Wow this new one is a slow sucker.
<hr /></blockquote>Darn, I just did a quick benchmark for the new protein and it looks like my 2*1GHz is on line to crunching ~28K/day from a previous ~80K/day. And the 500 MHz iBook is doing about ~6.5K/day compared to a previous ~16.5K/day. I'll know more accurately after 24hours.

I watched the iBook draw the ascii proteins for a bit to verify it was working before I restarted it in quiet mode. It was painful. It was chugging through the calculations. Sorry little iBook.

Shaktai · May 31, 2002, 03:22 AM

<blockquote>quote:<hr />Originally posted by enola:
 <blockquote>quote:<hr />Originally posted by Scotttheking:

Wow this new one is a slow sucker.
<hr /></blockquote>I watched the iBook draw the ascii proteins for a bit to verify it was working before I restarted it in quiet mode. It was painful. It was chugging through the calculations. Sorry little iBook.<hr /></blockquote>Yeah, but at least you know that little iBook will keep a-chuggin, even when some of the bigger dogs are plum wore out.

SkiBikeSki · May 31, 2002, 03:27 PM

Guess I should stop collecting benchmarks until after CASP, and the regular proteins return. Or maybe I'll start a new table...

jbcool · May 31, 2002, 04:18 PM

Glad to hear that the transition went well for most of you.
Came home twelve hours later and the only thing that updated were the linux machines. Mac OSX and Windoze XP machines didn't update.
So much for having the client set up to auto update.

But on the brighter side of things it did allow me to take the #2 Team MacNN spot at folding @ home while the distributed folding clents were at a stand still.

Good news bad news type of thing.
Ok Shaktai your turn to pass me now. <img border="0" title="" alt="[Wink]" src="wink.gif" />

serpico483 · May 31, 2002, 05:14 PM

no dice on the AUTO update , it tryed but failed.... deleted the old folders and down loaded the new..... seemed less painfull.....
but *amn it is SLOW

Scotttheking · May 31, 2002, 07:52 PM

linux updated, OSX didn't, dunno about windows, I don't feel like checking.

Shaktai · May 31, 2002, 10:05 PM

<blockquote>quote:<hr />Originally posted by Scotttheking:
linux updated, OSX didn't, dunno about windows, I don't feel like checking.<hr /></blockquote>All mine updated without a problem.

jbcool · May 31, 2002, 10:34 PM

<blockquote>quote:<hr />Originally posted by Shaktai:
 <blockquote>quote:<hr />Originally posted by Scotttheking:
linux updated, OSX didn't, dunno about windows, I don't feel like checking.<hr /></blockquote>All mine updated without a problem.<hr /></blockquote>Guess you know how to type a "1" better than the rest of us.

Glad to hear it went well for you.

mikkyo · May 31, 2002, 11:59 PM

<blockquote>quote:<hr />Originally posted by Shaktai:
All mine updated without a problem.<hr /></blockquote>As did mine, OS X and Linux.

Shaktai · Jun 1, 2002, 12:37 AM

<blockquote>quote:<hr />Originally posted by jbcool:
 <blockquote>quote:<hr />Originally posted by Shaktai:
 <blockquote>quote:<hr />Originally posted by Scotttheking:
linux updated, OSX didn't, dunno about windows, I don't feel like checking.<hr /></blockquote>All mine updated without a problem.<hr /></blockquote>Guess you know how to type a "1" better than the rest of us.

Glad to hear it went well for you.<hr /></blockquote>I made certain I proofread it well. I used an old proofreaders technique of reading it backwards. <img border="0" title="" alt="[Wink]" src="wink.gif" />

Last update I had some problems because I misnamed the autoupdate.cfg file. This time I made sure everything was perfect. Mac, Linux and Win machines all updated flawlessly.

jbcool · Jun 1, 2002, 11:46 PM

<blockquote>quote:<hr />Originally posted by Shaktai:


Last update I had some problems because I misnamed the autoupdate.cfg file. This time I made sure everything was perfect. Mac, Linux and Win machines all updated flawlessly.

<hr /></blockquote>I believe I might need to double check my work then.
Like I always say "Darn users messed things up again."

To error is human, for me it's an hourly occurrence. <img border="0" alt="[Hmmm]" title="" src="graemlins/hmmm.gif" />

Shaktai · Jun 2, 2002, 12:35 AM

<blockquote>quote:<hr />Originally posted by jbcool:
I believe I might need to double check my work then.
Like I always say "Darn users messed things up again."

To error is human, for me it's an hourly occurrence. <img border="0" alt="[Hmmm]" title="" src="graemlins/hmmm.gif" /> <hr /></blockquote>Sounds like that CRS acting up again. Ya think protein folding might help find a cure for that???????

jbcool · Jun 2, 2002, 09:51 AM

<blockquote>quote:<hr />Originally posted by Shaktai:
QUOTE]Sounds like that CRS acting up again. Ya think protein folding might help find a cure for that???????

<hr /></blockquote>CRS, Old timers disease, etc..................
Make sure that you are not talking with medical people when you mention old timers disease. They will try correcting you thinking that you don't know how pronounce the disease properly. Then you have to step back and explain what your talking about.

Scotttheking · Jun 2, 2002, 12:17 PM

all machines updated and running.
Now I just need a way to run it at bootup or cron, but I'm not in the mood, nor do I have the time, to get it working.
I even modified the sis's imac so it doesn't sleep anymore, hopefully she won't notice

enola · Jun 2, 2002, 12:47 PM

<blockquote>quote:<hr />Originally posted by Scotttheking:
I even modified the sis's imac so it doesn't sleep anymore, hopefully she won't notice

<hr /></blockquote>Ha ha. Then it's a good thing the iMac's don't have fans.

Scotttheking · Jun 2, 2002, 01:12 PM

<blockquote>quote:<hr />Originally posted by enola:
 <blockquote>quote:<hr />Originally posted by Scotttheking:
I even modified the sis's imac so it doesn't sleep anymore, hopefully she won't notice

<hr /></blockquote>Ha ha. Then it's a good thing the iMac's don't have fans.<hr /></blockquote>G4 imacs do have a fan <img border="0" title="" alt="[Frown]" src="frown.gif" />

wheeles · Jun 2, 2002, 01:44 PM

<blockquote>quote:<hr />Originally posted by Scotttheking:

Now I just need a way to run it at bootup or cron<hr /></blockquote>I've been using <a href="http://www.versiontracker.com/moreinfo.fcgi?id=14315&db=mac" target="_blank">LIOService</a> for running my client startup script on bootup. It also runs a shutdown script for the client on logoff too. I had to use this as I always forgot to manually shut it down when I turned the box off and kept losing work.