Dr. Baker's journal archive 2008

Message boards : Rosetta@home Science : Dr. Baker's journal archive 2008

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 51068 - Posted: 29 Jan 2008, 5:43:30 UTC

A lot of progress has been made in the last month!!

Our efforts to design new enzymes have been very successful, and it is likely you will be able to read about this progress in magazines available at local news stands--I'll keep you posted on this.

The new version of Rosetta, with many new capabilities, now appears to be stably running on Ralph, and David Kim is going to set it up on Rosetta@home in the next day or two. While the science will be new, the graphics will be a bit primitive at first, so please bear with us! We will be testing out new approaches to protein structure prediction in preparation for CASP8 this summer.

The rosetta game continues to improve with new features and new challenges. Until we are satisfied that all the bugs are out, we will remain in a "viral" distribution mode; if you would like to play you can request an invite from one of the volunteers testing out the current version.
ID: 51068 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 51408 - Posted: 15 Feb 2008, 4:01:17 UTC

These are exciting times!

first, please bear with us! the all new beautiful completely object oriented C++ rewrite version of Rosetta is now running on your computers. old timers will recall that in the early days of the project, it took sometime before all the kinks were worked out, and that the graphics took some time to get to the very nice state we have all become used to. the new version of rosetta is extremely powerful, but as it is very new, there are still some rare problems surfacing, and some of the bells and whistles, such as the screensaver, are not yet in place. similar issues have been slowing us down a bit with our in house research and computing projects. rest assured however, that these problems will be solved quite soon, and the new version is so much more powerful that this inconvenience will be made up for many fold.

second, the game continue to progress, and I think is getting more and more fun to play. still some things to work out, so not quite ready to go public, but I encourage any of you who are interested to give it a try (see thread below).

third, as I alluded to in my last post, we have two very exciting papers describing a major breakthrough--the computer based design of new enzymes catalyzing novel chemical reactions-- that will be appearing soon, for a preview you can look at the current online issue of Nature.
ID: 51408 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 51832 - Posted: 7 Mar 2008, 6:26:50 UTC

The most recent of the many scientific papers that have come out of all of your efforts at rosetta@home is in today's issue of Science magazine:

http://www.sciencemag.org/cgi/content/abstract/319/5868/1387

This paper describes the computational design of new catalysts for chemical reactions. Among the many applications are removal of toxic compounds in the environment (bioremediation), making much easier difficult steps in the synthesis of drug molecules, and the generation of new biofuels.

Thanks again for all of your help!
ID: 51832 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 52032 - Posted: 20 Mar 2008, 4:44:11 UTC

The 2nd rosetta@home enzyme design paper is now out in the online edition of Nature magazine:

http://www.nature.com/nature/journal/vaop/ncurrent/index.html

ID: 52032 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 52240 - Posted: 4 Apr 2008, 20:47:47 UTC

We are now trying to get the new version of rosetta up to speed (and more!) in time for CASP8 which starts in about a month. we will need to test new versions more rapidly than usual, and to make this possible without disrupting rosetta@home we would like to grow our test server RALPH. so if you are willing to tolerate test level code (which has a slightly higher frequency of crashing) please sign up on RALPH (we will be updating the executable as we make changes a couple of times a week so you will probably want to have a reasonably fast network connection). if you already split between rosetta and ralph, it would be helpful to increase your ralph time allotment. these adjustments are for the next month only--after this we should have a great new improved version of rosetta for rosetta@home that will be used for the casp8 targets. thanks! David
ID: 52240 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 52608 - Posted: 19 Apr 2008, 5:03:53 UTC

The Rosetta game is about to go public! beat the crowds and start working your way up the world rankings, while honing skills you can soon put to use in CASP and in designing new proteins to help people worldwide.

Download the game at

http://fold.it/boinc

Have fun!
ID: 52608 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 52767 - Posted: 28 Apr 2008, 3:33:05 UTC

David Kim has updated the "articles about rosetta and rosetta@home" page:
https://boinc.bakerlab.org/rosetta/rah_articles.php

if there are other articles out there that we have missed, please let us know and we will add them to the list.

again--the rosetta game will be going public in the next couple of weeks (hopefully you will be able to read about it in your favorite newspaper), so try it out now and beat the crowds!


ID: 52767 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 52902 - Posted: 7 May 2008, 19:10:15 UTC

CASP8 has started! Look for CASP8 targets to start appearing on your screensavers soon!

Again--your contributions are invaluable for testing the latest in methods for protein structure prediction. The results of CASP7 were published in several scientific papers, including an article in Nature, and we always cite the participants who found the lowest energy models. More generally, the results of your calculations will inform the world of what is currently possible with the latest in protein structure prediction.

The rosetta game "foldit" goes public on thursday. hopefully you will see something in your favorite newspaper; in any event, sign in now to beat the crowds! The game has improved a lot in the last several weeks, so if you haven't played in a while, give it another try.

David Kim has pulled off another amazing technical feat--our automated structure prediction pipeline Robetta will now be using rosetta@home for time intensive "abrelax" calculations. this will be the first time that robetta has been able to carry out the high resolution refinement that is necessary for improved accuracy. the results will be of particularly broad interest because a completely automatic protocol, if successful, can be run on many different proteins by many different groups. Since all of the rosetta code is freely available, institutions interested in particular sets of proteins can use robetta, if they want to depending on your results in CASP8, to predict their structures.




ID: 52902 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 52963 - Posted: 10 May 2008, 5:08:22 UTC

The rosetta game Fold It went live yesterday, and apparently 9000 people had signed up by early afternoon today. it has been very exciting to watch! the competition is getting fierce, and my son is complaining (correctly) that all my suggestions are making his score worse rather than better! we are keeping our fingers crossed about our servers ever since fold it made it onto slash dot yesterday afternoon; so far we have been able to keep running most of the time although there have been some slowdowns.



ID: 52963 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 53002 - Posted: 12 May 2008, 6:04:37 UTC

CASP8 is ramping up with a handful of targets released last week, and many more expected this week. The reason you haven't seen many CASP8 work units as we are making a final benchmarking of all the new structure prediction protocols we have developed over the past two years in the new miniRosetta code framework. The testing is on CASP7 targets--we want to determine which of our new methods are best before going all out on the CASP8 targets (yes, we would like to have completed the benchmarking before casp8 started, but there has been so much to do!). the testing should be complete monday and tuesday, and based on these results we will be sending out primarly casp8 work units next week using the approaches the tests currently running show are most successful.

I can't emphasize enough how critical all of your contributions are to our efforts to improve protein structure prediction methods! we have had many ideas on potential improvements which we have implemented, but the only way to know whether we are actually getting closer to the truth is to do rigorous testing (the casp7 targets for which we now know the structures are a perfect test set), and this would quite simply be impossible without your contributions!
ID: 53002 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 53192 - Posted: 20 May 2008, 4:59:56 UTC

James Thompson and Andrew Leaver-Fay through some heroic work this weekend (particularly given the beautiful weather in Seattle which has been all too rare this spring) tracked down the problem that was interfering with the new miniRosetta code running well on some people's computers and preventing us from launching large scale miniRosetta calculations on the CASP8 targets which have come out so far. the problem turned out to be not in the miniRosetta code but in the compiler, so it was especially hard to track down. With this fixed, you should see your computers running miniRosetta calculations on CASP targets very soon!
ID: 53192 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 53373 - Posted: 27 May 2008, 6:59:05 UTC

The first CASP8 target predictions were submitted today! There will be targets due every day or two from now until the end of July--keep crunching! As CASP7 veterans will know, we will all learn how well we did when the CASP experiment is assessed this fall.
ID: 53373 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 53581 - Posted: 7 Jun 2008, 19:27:59 UTC

CASP8 is now in full swing. We have prediction challenges due almost every day for the next couple of months. Keep up the great computing work, and please try to recruit your friends and relations--we need all the help we can get! Thanks! David


ID: 53581 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 53890 - Posted: 21 Jun 2008, 6:23:29 UTC

CASP8 is just over half way through, and it is proving to be a real challenge to keep up with all of the targets which keep streaming in. We apologize for the work units released earlier this week that caused problems--these were completely unanticipated based on the many similar calculations we have been doing. CASP is stimulating many new ideas for improving structure prediction methodology which we are trying to test out in parallel with completing the casp challenges, hence you may see some non casp work units alongside the casp proteins. Thanks again for your contributions!
ID: 53890 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 54076 - Posted: 30 Jun 2008, 5:49:46 UTC

As you have probably noticed, many of the CASP8 targets have been quite large proteins. In the last few days, around 5 new targets were released all below 150amino acids--we are very excited about the possibility of predicting these structures very accurately, and expect to see many work units on these proteins in the next several weeks!.
ID: 54076 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 54682 - Posted: 27 Jul 2008, 4:21:29 UTC

we are down to the last week or so of CASP8! one great thing about casp this year is the many new ideas that have come up during the process, and we are all excited to test these ideas systematically once casp is over. so be ready for a flurry of testing of new methods in August--I'm optimistic this will lead to significant improvements in protein structure prediction!
ID: 54682 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 55497 - Posted: 3 Sep 2008, 5:55:49 UTC

CASP8 is now finally over! After the "normal" prediction season, we were sent a new type of challenge--to refine a starting structure model to increase its accuracy and decrease its energy. These refinement challenges occupied us for much of August.

The structures of many of the CASP8 targets have now been made public and we are comparing the predictions made on your computers to the actual structures. There are already some very nice predictions in the set we know the answer for so far!

Those of you who particpated in CASP7 will remember that it is not until the CASP meeting in December that we can compare your results to those of other groups. We of course hope the results will be as good as in previous CASP challenges!

In other rosetta@home news, you will have noticed that David Kim has done an amazing job with the mini-rosetta graphics as it now is as beautiful as the old rosetta graphics was--I hope you are enjoying looking at it.

Our goal for the next months before the CASP meeting is to automate all of the new techniques we developed during casp8 to create an automatic pipeline that can generate protein structure models as accurate as those we generated in CASP8. Needless to say, this will require a very large amount of testing and benchmarking of the new methods and you will see these jobs running on your computers soon!

Thanks again for your contributions during CASP8 and during this next period where we try to push the envelope of protein structure prediction methodology!
ID: 55497 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 55707 - Posted: 12 Sep 2008, 3:53:25 UTC

Reporter John Bohannon is writing an article for WIRED on rosetta in CASP8 and would like to interview rosetta@home participants who contributed to CASP8 structure predictions about their experiences computing this summer. He can be contacted through

https://boinc.bakerlab.org/rosetta/forum_thread.php?id=4367&nowrap=true#55703
ID: 55707 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 55730 - Posted: 13 Sep 2008, 6:03:20 UTC

The official results from CASP8 will be made public at the CASP8 meeting in early December. For those of you who are interested in how models are evaluated and the current problems in structure prediction the official assessment paper for the hardest category of targets in CASP7 is posted at

https://boinc.bakerlab.org/rosetta/rah_misc/casp7_fm_assessment.pdf

ID: 55730 · Rating: 0 · rate: Rate + / Rate - Report as offensive
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 55826 - Posted: 17 Sep 2008, 5:25:17 UTC

John Bohannan from WIRED will be finishing up his article soon, and so far has only heard from two rosetta@home participants. Please respond to his post on the Cafe Rosetta boards soon if you have opinions/thoughts you would like to see in WIRED.
ID: 55826 · Rating: 0 · rate: Rate + / Rate - Report as offensive
1 · 2 · Next

Message boards : Rosetta@home Science : Dr. Baker's journal archive 2008



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