DF Daily (01/22/03)

[m0ti]

Hi Everyone,

Well, today the RMS jumped by a hop, skip and a jump, and by leaps and bounds at that (well, only one leap, but it was a big one). The best RMS is now at 6.53. Anybody notice since we moved to the new algorithm that the best RMS tends to really jump every once in a while?

ATWE
as of: Wed Jan 22 16:10:31 2003 EST
GhettoFob gets 323,057 ticket 1-22-03 0300

Position * Contestant * Current Folds - Starting Folds - Tickets = ATWE Total

1 * ATWE Terrahawk * 159,028,036 - 465,147 - 0 = 158,562,889
2 * ATWE Robor * 173,834,553 - 38,230,740 - 15,820,786 = 119,783,027
3 * ATWE LastKnight * 78,490,972 - 3,483,399 - 0 = 75,007,573
4 * ATWE Insidious * 131,578,563 - 58,684,902 - 681,901 = 72,211,760
5 * ATWE ReaganCRW * 65,811,901 - 10,501,161 - 0 = 55,310,740
6 * ATWE bfour * 78,697,471 - 22,005,703 - 3,077,288 = 53,614,480
7 * ATWE cbcrus * 67,886,783 - 28,117,719 - 0 = 39,769,064
8 * ATWE bob970 * 28,496,420 - 1,243,892 - 94,396 = 27,158,132
9 * ATWE FifthFactor * 26,109,625 - 684,121 - 120,000 = 25,305,504
10 * ATWE bunker * 25,478,551 - 830,790 - 958,530 = 23,689,231
11 * ATWE m0ti * 63,563,277 - 40,222,221 - 0 = 23,341,056
12 * ATWE muttley * 58,008,343 - 38,759,122 - 540,880 = 18,708,341
13 * ATWE GhettoFob * 21,233,749 - 6,169,633 - 574,898 = 14,489,218
14 * ATWE CrazySaint * 16,877,586 - 2,857,131 - 724,626 = 13,295,829
15 * ATWE Builderz * 15,251,446 - 2,429,825 - 0 = 12,821,621
16 * ATWE joinT * 13,196,487 - 63,313 - 922,653 = 12,210,521
17 * ATWE Mloot * 18,881,708 - 7,389,298 - 0 = 11,492,410
18 * ATWE TwilightZone77 * 25,689,587 - 15,566,340 - 0 = 10,123,247
19 * ATWE ken008 * 7,186,346 - 1,898,137 - 487,351 = 4,800,858
20 * ATWE Rebel Alliance * 4,475,000 - 5,000 - 80,000 = 4,390,000

Milestone Makers

ATWE bunker hits 25M folds!

😎
ATWE Builderz hits 15M folds!

😎
ATWE ken008 hits 7M folds! 😎😀😀

Congrats to the milestone makers! 🙂😎😀

FoBoT: A better day than yesterday, production-wise. 10.4M folds and 3rd overall! RMS took a tumble today, so you'll either have to make an awesome fold or be content with 2nd...

non-FoBoT:
Top Folder: ATWE Robor with 173.8M folds and 68th! 😀
Best RMS: ATWE Terrahawk with 7.10 and 6th! 😎
Best Production: ATWE Robor with 4.2M folds and 19th overall! 😉

Producers Of the Day

FoBoT 10.4M 😛
ATWE Robor 4.3M 😛
ATWE Terrahawk 3.8M 😛
ATWE bfour 2.2M 😛
ATWE LastKnight 2.1M 😛
ATWE Insidious 1.8M 😛
TA Rescue Team 1.4M 😛
ATWE ReaganCRW 1.4M 😛
ATWE cbcrus 1.2M 😛

Congrats to the POD's! 😛😛😛

TeAm Stuff

Another fine day, for TADF. 55% user activity ain't bad. It's looking more and more likely that the protein change will be on Tuesday. As a reminder to Terrahawk, whom I don't know if he saw my response from a couple of days ago, you can use dfDetect to run under Win98 as a "service" (though it MAY not exit out as cleanly as a normal service would... it does try though and it shouldn't pester the user).

Daily Production: 35.8M and a 2nd place showing once again for the TeAm! Awesome stuff, guys! I think that's our 3rd day in a row or so. We've also got the 2nd strongest showing for the weekly production! 😎

Great stuff! See y'all tomorrow!!! 😀

 

[Insidious]

Thanks m0ti!

Another great day for the TADFers! 😎

-Sid

 

[FoBoT]

there is some news on the DF site

01/22/2003

- We are fast approaching 10 billion again, and have a 6.5A structure! So congratulations to all. This particular protein was expected to get roughly a 6A structure in a sample of 10 billion so we have come very close to that prediction. We will be changing proteins again next Tuesday afternoon, EST. Also coming soon, a brand new smarter algorithm. Watch here and in the discussion forum for details.

- Results from the CASP5 structure prediction competition have been posted in the results section, with a brief explanation of how to interpret the graphs. In most cases our predictions were in the middle of the pack. This might not sound as good as we had hoped, but it means none of our predictions were THAT bad, and in general all five of our top predictions were of comparable quality, meaning that we were picking out structures consistently. It is also important to keep in mind that of the 13 targets predicted, only T0129, T0139, T0161 and T0181 were truly 'new folds'. Other targets could in some cases be predicted by other means starting from related known structures (though still by no means easy to predict). We still need to go through the structures we generated to see if we submitted the best predictions we could have, or if there were better structures that our scoring function simply failed to pick out. Clearly brute force alone is not sufficient to beat other more intelligent search methods though, and we will shortly be revising the algorithm to work in a more intelligent, iterative sampling approach, which we expect to have a huge effect on the overall performance and efficiency of the sampling.

 

[GhettoFob]

GhettoFob gets 323,057 ticket 1-22-03 0300

😱 Thanks m0ti!

 

[RU482]

anyone know if , on our taxes, we can write off the computer portion of our electric bill ?? Dammit, we need a lobbyist in Congress!!

 

[Terrahawk]

Thanks for the stats, m0ti!

As it turned out, I didn't see your post the other day but have now downloaded dfdetect and will install it with the machines...
once the RAM gets here! It turned out that the package I got was not the RAM, and there has been an shortage of SDRAM at our supplier!

Ah well, I'll keep hassling them.

Congratulations to the Milestone Makers and the Producers of the day! Keep up the good work!

 

[m0ti]

hmm, taking a look at the results we do seem to be in the middle of the pack, sometimes in the high-end, sometimes in the low-end, ...

one thing howard forgot to mention though, was that for most of those proteins we only did about 1 billion folds (I think). I think that between the new algorithm we're using now, which is faster (2X) and produces better results, and the new smarter algorithm on the way that we could get much much better predictions.

I hope that we start doing the CASP 5 proteins soon, instead of those from CASP 4, and improve the results we previously had.

 

[Insidious]

Thanks FoBoT!

or if there were better structures that our scoring function simply failed to pick out.

Hmmmm, I think maybe Howard is beginning to see what you and I did m0ti! At least, he has realized it
is worth a second look. And on that subject, the RMS plot shows a decrease in the number of Best RMS
values that changed for us as the protien goes on. The next couple days of plot might be interesting.

-Sid

(Comcrap won't let me login at the moment, I'll get the latest uploaded ASAP)

 

[m0ti]

Sid,

I think howard is talking about inherent limitations in the scoring functions. The scoring functions provide a quick and rough estimation of how good a particular fold is; hopefully, the scoring functions are fairly accurate in maintaining the order between folds (i.e. if score(a) > score(b) then a is in fact better than b, even when checked with more exact scoring functions). It is possible that the scoring functions are not grading certain categories of folds correctly.

What we are talking about seems to be something with the random number generator or similar affects.

once we switch the protein over I'll compile some stats looking at number of folds produced vs rms improvements as expected etc.

 

[Robor]

/me is confused on all this tech talk on the project. 😕 😱 I have no idea what you guys are talking about. I'll just keep the machines going and try and keep my numbers up. 🙂 I added a P3-650/256MB last night. I'm building it for a friend of my mom and I'm using DF to test the stability. 😀

😀 Congrats to the POD's & MM'ers and thanks m0ti & Sid! 😀