| Methods abstracts | (2010-8-30) |
Dear CASP9 predictors,
It's abstracts time! The Abstract submission web page is available through the link from the CASP9 main page. Please read the instructions before submitting your abstract as format requirements have changed. The submission deadline is September 20, 2010. Please remember that your contributions will be taken into account in choosing some presentations at the meeting. Also, the predictors presenting posters at the meeting should be prepared to give a short presentation at one of the main sessions as some talks will be invited during the meeting based on the discussion of poster sessions.
Short update on the CASP9 evaluation progress. At this moment, the Prediction Center team and assessors are working at full load evaluating your predictions. As of today, the PDB have released structures for 78 CASP9 targets. You can check which targets have been already released by the PDB from our target list page. We also have more than a dozen structures at hand received directly from the experimentalists. The preliminary evaluations have been run on all of these structures. The assessors are at the final stages of defining domains and categorizing the targets so that the final evaluations can be compiled.
|
|
| End of the CASP9 prediction season | (2010-7-30) |
Tomorrow is the deadline for submitting predictions for the remaining 3 out of 4 human/server CASP targets (T0643 expires next week). The final CASP9 target statistics: 129 targets including 60 human/server targets. We expect the majority of CASP9 structures to be gradually released by the PDB in the next 5 weeks. We will be updating our target list weekly (on Wednesdays). As of today, thirty-nine CASP9 targets have already been released by the PDB.
Also, today we have released the last (and we believe the most challenging one) target for the refinement. The details of the suggested target are as usually provided in the "Additional information" field of the target page. All in all, we have released 14 targets for the refinement experiment in CASP9. This is two targets more than we had in CASP8.
Now it's a good time to thank to all of you who made CASP9 yet another successful protein structure prediction experiment. We have already received a record number of 85,000 (!) predictions from you this season, and still counting. Thanks to your dedication and the hard work! Now it's time for our computers at the Prediction Center and the assessors' teams to evaluate the predictions. We'll keep you posted...
CASP9 organizers
|
|
| T0622 - additional information | (2010-7-25) |
|
T0622 expires for human prediction coming Tuesday. Those, who participate in the refinement experiment should have noticed the additional information we have posted about this target on the Refinement Target List page. We want to make sure that all the predictors are in equal conditions and noticed this info.
|
|
| TR614 and TR622 | (2010-7-23) |
We released two more refinement targets today. Please note that even though we suggested two different models as starting points for TR614 refinement, it is technically only one refinement target and only one structure
should be provided in the prediction (no residue repetitions).
Next week we will conclude the release of the refinement targets. Hoping for 2-3 more targets depending on structure availability and suitability of the models for the refinement.
|
|
| T0637 - reverted to server only target | (2010-7-21) |
|
Target T0637 was independently solved by another crystallographic group and publicly released by the PDB today. We are reverting it back to the server only status.
|
|
| July 17 - the last day of the target release (update - 2) | (2010-7-17) |
In the first two hours after releasing today's targets we have received a couple of requests to extend shorter than usual deadlines for the human/server targets released this week. We have discussed these requests and decided to slightly spill over into the first week of August in order to give you the full 3 weeks for prediction of T0637 and T0643.
CASP9 organizers
|
|
| July 17 - the last day of the target release (update) | (2010-7-17) |
Just a few hours ago we received from our target depositors a sequence that shows no significant similarity to sequences of the PDB proteins. As hard prediction targets are getting more and more rare, it would be pity to waste such a good target. Therefore, in addition to the two server targets prepared for today's final release, we will be releasing another human/server target - T0643.
Sorry for the somewhat hectic last week of the target release season.
|
|
| July 17 - the last day of the target release | (2010-7-16) |
Dear CASP9 participants,
As of today, July 16, we have released 126 targets for prediction including 61 in the human/server category. Please pay attention that the yesterday's target T0637 has been switched to the human/server category to compensate for the target reverted to the server only category yesterday. The last additions to the target list will be two more server targets that are scheduled for release tomorrow. And that would be it for regular prediction targets!
The CASP9 target statistics figures are well in accordance with the plan that we had for this round of CASP and very similar to the CASP8 figures (128/57). We wish to thank all the researches who were part of the CASP9 target collection effort, especially to the crystallographers and spectroscopists from the 3 Structural Genomics Centers (JCSG, NESG and MCSG), who provided more than 100 targets altogether for the CASP9 experiment.
We will continue pushing tasks to the QA servers until the last server target expires for prediction (July 21).
One more model was added to the refinement experiment today (TR606) and we expect to add a few more within the next two weeks.
|
|
| T0555 is reverted to the server only targets | (2010-7-15) |
|
Yesterday we announced about moving T0555 to the human/server category. On the same day, unluckily for us, the PDB released structure of 2ky4, which is our another target - T0553 and a close homologue of T0555. Under these circumstances prediction of T0555 by human-expert groups is not that challenging any more (for those who update their structural libraries regularly, of course). Therefore, we return T0555 back to the server only status and no new predictions will be accepted on this target.
|
|
| Switching T0555 and T0562 to human/server prediction category | (2010-7-14) |
|
Two more targets released in the "server only" prediction track, T0555 and T0562, appeared to be hard enough for modeling. We want to check if human-expert groups can do better on these targets and shifting them to the human/server prediction category. The human prediction deadlines for these targets are reset to July 31.
|
|
| T0600 canceled for human predictions | (2010-7-12) |
We have canceled T0600 for human prediction as ribbon diagram of the structure had been prematurely exposed on the Internet. The target has been moved to the "server only" prediction category. Server predictions only will be assessed for this target.
Two more targets were released for the refinement today. We may potentially release one or two more later this week.
|
|
| Week of July 12 - the last week of CASP9 target release | (2010-7-9) |
|
We have almost finished releasing human/server targets for prediction in CASP9. We are at the number 57 so far. This week we released more targets of this type than usual in order to give you full 3 weeks to predict each of them. Next week is the last, 11th week of CASP9 target release and we will be releasing predominantly "server only" targets (based on their availability) with a shorter human deadline for them. If something really, really interesting shows up during this last week and it will be pity to waste such a target for server only prediction - we may release it as a human/server target. There will be at least one such a target released next Monday, but not more than 3 of them for the whole week.
|
|
| Another starting model for TR517 and new refinement targets | (2010-6-28) |
While predictors can use any of the server models as the starting model for the refinement, we advise you which models turn to be the best so that the refinement can work better for you. Last week we released TR517 for refinement selecting one of the best server models as the starting model. This model had some missing residues while the best human model had all the residues in place. We want to provide you this model as a potentially better starting point for the refinement. Please recheck the Refinement Target List page for the new starting model TR517.
We have also released 3 new refinement targets and are planning on releasing another one tomorrow.
CASP organizers
|
|
| CASP9 update - June 25 | (2010-6-25) |
Only three weeks are left for the release of targets this season. So, if someone knows a crystallographer or a NMR-spectroscopist who might have interesting targets for our experiment, please ask them to submit those to CASP right away.
This week we have released the first two targets for model refinement - TR517 and TR530
(http://predictioncenter.org/casp9/targetlist.cgi?view=refinement). The details of the suggested refinement targets are provided in the "Additional information" field for the targets. Coordinates of the starting models are available from the field "Template". Please, submit your refined models the same way you do for the regular predictions, remembering that names for the refinement targets start with "TR" instead of "T0".
Next week we are planning on adding several more targets to the refinement experiment and 10 targets for regular predictions.
CASP organizers
|
|
| T0581 category change and Refinement targets | (2010-6-23) |
We originally released target T0581 as a server only target. After more careful consideration it appeared that this target is one of the more challenging targets to predict. We would like to suggest this target for prediction to human groups and shift it to human/server prediction category. To allow extra time for expert prediction, we are extending human deadline for T0581 by July 11.
We also will be releasing our first refinement targets this week. Please check our Refinement Target List link from the main CASP9 web page.
|
|
| CASP9 update - May 14 | (2010-5-14) |
In our previous communication we stressed that human-expert structure predictors are recommended to concentrate their efforts on modeling of Human/Server targets. We want to emphasize here that all groups (including human-expert) that plan to predict in DR/RR/FN/QA categories are expected to submit their predictions on all targets, i.e. both - "Human/Server" and "Server only".
Disorder region predictors have been informed by the organizers that we had started enforcing the rule
that the residues predicted to be in the disordered state should have probability scores >0.5, while those in ordered state - less than 0.5. For those who are still planning to submit DR predictions - please make sure that your format comply with the requirements posted at our website.
Next Monday we will release one big target (887 residues) that will be followed by 10 to 12 other targets throughout the week.
CASP9 organizers
|
|
| CASP update - May 7 | (2010-5-7) |
First week of CASP9 prediction season is over. We have released 14 targets. The vast majority of them were easy TBM targets. Next week you will find some harder targets in the human prediction category.
As of today, we have 125 groups (predominantly servers at this early stage) contributing models to the Prediction Center. You can always find the latest CASP statistics at http://predictioncenter.org/casp9/numbers.cgi .
In CASP9, we are planning on releasing between 50 and 60 targets for human-and-server prediction and approximately twice as much targets (depending on their availability) for server-only prediction. Assessors will be comparing all participating predictor groups on the subset of Human/Server targets and, additionally, the server groups on all released targets. The human-expert groups should concentrate on predicting structures for the Human/Server targets. Those, wishing to take the challenge and predict all of the released targets (including the "server-only" ones) are welcome to do so but the evaluation emphasis will be placed on the "Human/Server" targets. Subject to availability, we will give priority to releasing targets containing low homology domains in the Human/Server track.
Reminding that we are accepting multimeric predictions this CASP. So far, we received such predictions from 9 servers.
Tarballs of tertiary structure models from servers are made publicly available at http://predictioncenter.org/download_area/CASP9/server_predictions/ the next day after a particular target is closed for server prediction, i.e. on the forth day after the target release, at 7:45am PDT. Approximately an hour later these tarballs are submitted to the quality assessment (QA) servers, which have an extra 3 days to submit their QA predictions. Deadline for human-expert QA predictions is the same as for all other types of prediction.
On the weekend days we will be submitting subsequent targets to the QA servers.
Next Monday we will release 2 new targets for prediction in all prediction categories.
Have a productive next week of CASP!
CASP organizers
|
|
| May 3 - CASP9 starts | (2010-4-29) |
We are happy to inform you that as of today, April 29, we already have 100 human-expert groups and 125 automatic servers registered for the CASP9 experiment. With practically all CASP8 "winners" already on board and around 20 "new kids on the block", we are anticipating another exiting prediction season. Structural Genomics centers have started submitting sequences for potential targets to us and we have already prepared enough targets to run the first 2 weeks of the experiment at full load.
We will kick off the 9th round of CASP experiments with two relatively easy targets next Monday, May 2. These two targets will be followed by three targets Tuesday through Friday totaling 14 targets for the first week. To start it gently, we will not be releasing very hard for prediction targets this week. The first difficult targets are expected to be released starting the second week of the prediction season.
Reminding you that according to the decision of the CASP8 predictors' meeting we are reviving quaternary structure prediction in CASP9. We will advise you on the possible oligomeric state of the particular target protein if such information is furnished to us by the experimentalists. We want to stress out that this information should be taken just as an advice on the MOST PROBABLE quaternary state of the target as data on many of the CASP targets are coming to us when structure refinement is still in progress and therefore experimentalists can not be sure of the final assignments. Format for the multi-chain predictions is described at the CASP9 format page:
http://predictioncenter.org/casp9/index.cgi?page=format#TS , and shown in the Example 4 there.
Andriy Kryshtafovych
for CASP9 organizers
|
|
| Second round of the dry run for servers | (2010-4-20) |
We have finished the first round of checking connectivity and format compatibility with the registered servers.
All non-QA servers were sent a query and we expected that all of them should have returned predictions by now (more than 3 days after the request). If your server did not get a query or did not send out a prediction - it's time to check for the problems and get in touch with the organizers. During the last 4 days we were able to identify and fix problems for the majority of 100+ servers registered for CASP9. Those server curators who haven't managed to make their servers work properly yet - don't hesitate to contact me in case you need help from the organizers' side.
Today we start a second round of the dry run for all registered CASP9 servers. In a few minutes we will be sending another test "target" to all non-QA servers. In the evening (Pacific Time) we will be sending a tarball from the first test "target" T0987 to the Quality Assesment servers.
Based on the first round of the dry run we can see that almost all of HTTP servers reply either with the web page or an email about their target receiving. This is not the case, though, for almost half of the email servers. So, email server curators: please, make sure that your servers reply to our distribution server casp8-meta@predictioncenter.org immediately after you have received a query. Please put the following text into the Subject of the email: "T0987 - query received by MY_SERVER". This will help us to track whether your server received a request from us so that we can timely address any connectivity issues. The prediction itself should be sent to the address specified in the REPLY-E-MAIL field of the query (please note that this address should be always taken from our query and not hard-coded as we may change it during the season).
Reminding you that during the dry run, you will be receiving both, acceptance and rejection messages from our verification server. During the regular CASP season it will be possible to check status of your server submissions through the link Server Predictions from the CASP9 web page: http://predictioncenter.org/casp9/status_n.cgi (link itself is already operational though not posted to the CASP9 web page yet).
Also, you can see all of your accepted predictions through the "My CASP9 profile" link from the CASP9 submenu on the left side of the CASP9 web page. The direct Model Viewer link will be available during the regular season from the CASP9 web page: http://predictioncenter.org/casp9/groups.cgi .
|
|
| Server dry run has started | (2010-4-16) |
At this, first stage of the dry run we are testing servers returning predictions in all formats but QA. Quality assessment servers will be tested starting next Tuesday.
During these tests you will be receiving both, acceptance and rejection messages from our verification server. During the regular season, though, acceptance messages will be turned off.
A request to all EMAIL-server curators: please, make sure that your servers reply to our distribution server casp8-meta@predictioncenter.org immediately after you have received a query. Please put the following text into the Subject of the email: "T0987 - query received by MY_SERVER". This will help us to track whether your server received a request from us so that we can timely address any connectivity issues. The prediction itself should be sent to the address specified in the field REPLY-E-MAIL of the query (please note that this address should be always taken from our query and not hard-coded as we may change it during the season).
|
|
| Server dry run reminder | (2010-4-14) |
We were notified by several server curators that their servers are still in the process of preparation for CASP9 dry run. To allow them some extra time, we are postponing testing connectivity with the servers for two days. Also, we advise those caspers who plan to participate in server track but have not registered their servers yet to do it immediately. On Friday, April 16, 2010 we will send a test target (T0987) to the registered servers. Please respond as you would for CASP9 targets. Of course, these predictions will not be part of the CASP9 experiment.
We will be sending 3 variables to your server's submission URL (or email): the SEQUENCE, the TARGET-NAME and the REPLY-E-MAIL (where to return the results). For the quality assessment servers we will be sending the TARBALL-LOCATION variable instead of (or in addition to, if you specify so) the SEQUENCE. Please, make sure you provide the required parameters in your CASP8 server registration.
IMPORTANT THINGS TO REMEMBER
There will be no essential changes in the procedures for submitting targets and accepting predictions comparatively to the CASP8. For CASP novices, helpful information can be found below in this announcement and at http://predictioncenter.org/casp9/registration.cgi , http://predictioncenter.org/casp9/index.cgi?page=format .
1. Target distribution procedure for servers.
You will be receiving CASP9 queries from the distribution server in Davis casp8-meta@predictioncenter.org and the results will be collected directly by the Prediction Center. At the end of each week we will try to notify you about how many targets we plan to release the following week. Targets are planned to be released on business days only. We plan not to exceed a load of 3 targets per day for servers. Targets will be made available and requests to all servers will be sent around 9am PST. The submission engine will resend the query if it encounters obvious connecting problems (network timeouts, 'no response' etc.). Failures that go beyond that require special attention, but we'll make every effort to notify server curators ASAP if we suspect something is not working. The facility that allows checking data flow status by server predictors will also be available.
2. Server submission format.
Please submit your predictions in the standard CASP format. Your submissions will be immediately verified at time of submission. CASP warning and error messages will return directly and immediately to you, and will refer to the material you submitted, allowing you to respond in close to a server time frame, if necessary. Please note that, as in previous experiments, no prediction will be finally accepted by CASP until it passes CASP format verification. Please, consult the CASP format description page for details:
http://predictioncenter.org/casp9/index.cgi?page=format
3. The timeframe for server prediction acceptance is the following. All server groups have 72 hours since target distribution to return the prediction. Notifications will be sent to the server contact person's email address only in case of errors encountered. If your prediction doesn't pass the CASP verification system, you will need to correct your format and manually resubmit the corrected prediction to the email servers@predictioncenter.org within the original 72-hour window. After 72 hours no server predictions will be accepted or altered in any way. If you expect your server to work on prediction generation more than 72 hours, then we would advise you to register a regular (not server) prediction group for CASP9 instead.
In addition to notification emails, you will be able to check the submission status of your prediction through the Server Submission Status Page or through the Model Viewer facility.
CASP9 organizers
|
|
| CASP9 - registration | (2010-3-29) |
The registration for CASP9 experiment is now open.
To register for the experiment, you will need to register with the Prediction Center (PC) first (if you haven't done this yet). At this step we will collect the basic registration information to be used for your participation in CASP9, all subsequent CASP experiments, between-CASP initiatives, and local services. The information provided by you will not be shared with anyone outside of the CASP system.
After you register with PC and login to our website, go to the CASP9 web page, where you will be able to register for the CASP9 experiment (a separate registration form under the CASP9 menu). You will find five separate registration forms there. Please, read the registration instructions and use the appropriate form.
We ask that you register with the Prediction Center at your earliest opportunity as we plan to release future CASP9-related news mainly through our website. We recommend that you select an option "Subscribe for Prediction Center newsletters" as in this case you will be receiving emails every time the news are posted on our website. This functionality was just recently added to our registration form. So, if you have registered earlier, please go to "My personal Data" link and select this option.
Also, we ask server group leaders to register as soon as possible as we are planning on testing the connectivity with the registered CASP9 servers starting April 14, 2010. The details of the dry run will be posted in the Message Board section of our website shortly (this section is available from both the Prediction Center home page and the CASP9 home page).
|
|
| Blue Waters computer | (2009-12-14) |
Dear CASP Participant:
You might like to know that NSF is soliciting proposals for developing applications for the Blue Waters computer, due to come on line in the next year. As you know, there is a lot more interest in compute intensive methods of modeling protein structure than there was a few years, and this is an opportunity for those of you who can see possibilities in another couple of orders of magnitude of compute power to get unique resources. NIGMS is involved in this because they are interesting in encouraging projects in the area of protein folding and structure modeling, among others.
There is a webcast this Thursday, at which you can get more specifics. The NIH announcement (From Peter Preusch is as follows:
"NIGMS seeks your input on areas of science that would benefit from access to the latest generation high-performance computer. The Blue Waters petascale computing system, which is under construction by the National Center for Supercomputing Applications (NCSA) at the University of Illinois and funded by the National Science Foundation, will be the most powerful computer in the world when it comes on line in 2011. Details of the capabilities and architecture of the Blue Waters computing system are available at http://ncsa.uiuc.edu/BlueWaters.
This message is to alert you to the opportunity to apply for allocations of computing time on the Blue Waters computing system and to help organize collaborative groups to submit high-impact community applications for time on this new machine.
NIGMS is facilitating the formation of collaborative groups by:
1) Hosting a virtual workshop and applicant briefing videocast, NIGMS-NSF Briefing on Blue Waters High-Performance Computing Opportunities, Thursday, December 17, 2009 at 2:00 - 4:00 p.m. on the opportunity to apply to NSF for allocations of computing time on this important new resource.
2) Providing a Web site where you can share ideas on areas of science that could benefit from access to this resource and express your interest in joining or leading a group of scientists in an application to the NSF.
For additional information, visit http://www.nigms.nih.gov/bluewaters."
John Moult.
|
|
| CASP9 meeting venue and dates | (2009-11-16) |
|
CASP9 predictors' meeting will take place at the Asilomar Conference Center, Pacific Grove, California on December 5-9, 2010. Details of the future CASP9 experiment are available from the Prediction Center website.
|
|
| CASP9 preliminary timetable | (2009-9-11) |
CASP9 experiment will be run May through July 2010.
Tentative dates are as follows:
March 30, 2010 - registration starts
April 14 - server "dry run" starts
May 3 - first targets will be released on or shortly after this date;
July 17 - last targets will be released not later than this date;
July 31 - prediction season ends;
December 2010 - the predictors meeting takes place at the Asilomar Conference Center in Pacific Grove, CA, USA.
CASP9 organizers
|
|
| CASP8 domain definition and coordinates | (2008-11-21) |
We have released the official domain definitions and coordinates used for CASP8 evaluations. CASP8 Abstract book is also available from our website. Results of the automatic evaluation of the predictions will be available one week before the meeting.
CASP8 organizers
|
|
| Arriving to Sardinia | (2008-11-16) |
A few important information about your travel to Sardinia:
Once you get to the airport of Cagliari, there are two ways to get to the hotel:
By bus: You take the ARST bus to Cagliari. the timetable can be found at http://www.sogaer.it/aeroporto/pullman/ . The cost is 2 Euro and the ticket can be bought at the automatic machines located in the hall of the airport.
The bus leaves you at the bus station in Cagliari. The station is on a square (Piazza Matteotti). You take the bus PF (direction Fiumini) and get out at the Hotel Setar stop (about 40 minutes). The ticket is 1 Euro.
You can take a taxi. The cost is about 40-50 Euro. Please only take official taxis and ask the driver the approximate cost to Hotel Setar.
There will probably be several people coming to CASP in your plane. You might want to share a taxi with them.
Meals included with your hotel room: from dinner on the day of arrival to lunch on day of departure. Any other arrangement should be made directly with the hotel. The cost of an extra lunch or dinner is 22 Euro.
The hotel bill should be settled upon arrival.
The hotel has wireless in all common areas and in most of the rooms.
|
|
| CASP8 job fair | (2008-10-28) |
Dear CASPers,
we would like to take advantage of the CASP8 meeting to help young scientists looking for a position in Computational Biology meeting potential group leaders and vice versa.
We plan to organize an afternoon to get them together for interviews / informal meetings /etc.
If you are looking for a position, could you please send your cv to anna.tramontano@uniroma1.it by the 9th of November?
If you have positions available could you please send a mail with the advert to anna.tramontano@uniroma1.it by the 9th of November?
As soon as we have an idea of how many people are interested in this initiative, we will get back to you with more information.
Thank you very much
The CASP organizers
|
|
| CASP8 junior scientist session | (2008-10-28) |
Dear CASP Participants,
During the CASP8 meeting we will dedicate a special session to topics
suggested by junior scholars. We will also award a poster prize and invite
oral presentations. All students, postdoctoral associates, and others young
at heart are welcome to participate (only junior scholars will be eligible
for awards ;-)).
Our goal is to select areas that are of particular interest to the
community of young scientists. Please send suggestions /issues to be
discussed to Krzysztof Fidelis (kfidelis@ucdavis.edu). We would also
welcome volunteers wishing to help organize this session (contact Krzysztof).
Krzysztof Fidelis
|
|
| CASP8 fellowships | (2008-10-15) |
We have now completed awarding CASP8 registration fee /lodging /travel fellowships and the recipients have been notified. The CASP8 speaker selection will be completed around Oct. 23. CASP8 speakers will be offered registration fee waivers, as well as lodging and travel reimbursements. At the same time we are extending early registration rate deadline to Oct. 27.
CASP8 organizers
|
|
| CASP8 meeting registration; December 3 - 7, 2008; Sardinia, Italy | (2008-9-18) |
Registration for the CASP8 meeting is now open. We expect that this year we will be able to accommodate all the people wishing to participate. The Meeting Registration Form is available at:
http://predictioncenter.org/casp8/meeting_registration.cgi .
Meeting fees consist of a registration fee and a lodging fee for four nights (December 3-7). Rough estimate of total meeting fees (full board) for an academic participant is $880-$1050 depending on the accommodation choice. Details are provided in the registration form.
A limited number of accommodation fellowships and travel fellowships will be available. However, registration should not be made contingent on an accommodation or travel award.
Predictors attending the meeting are also required to ensure that a long abstract and a questionnaire (if requested by assessors) describing their group's method have also been submitted. A web page for submitting your abstract is available at http://predictioncenter.org/casp8/abstracts.cgi .
Please note that meeting registration will not be complete without submitting these materials.
|
|
| Last 5 targets for refinement | (2008-8-2) |
|
Five targets for the refinement experiment have been released yesterday and today. The details of the suggested refinement targets are provided in the pdb files for the starting models and in the "Additional information" field for the targets. All in all, we released 12 targets (14 domains) for the refinement experiment in CASP8 .
|
|
| End of CASP8 target release | (2008-7-18) |
With the two server-only targets today, all CASP8 targets have been made available for prediction. In this round of CASP we have released 128 targets including 57 human/server targets. One target has been canceled so far. We expect the corresponding structures to be gradually released by the PDB. We regularly update our target list to indicate which targets have already appeared in the PDB. Please, check our target list for updates on Wednesdays. At this time we have structures for 34 CASP8 targets available.
Please note, that we will still be releasing targets for the refinement experiment. You may expect new refinement targets within the next two weeks. This week we have released 3 more targets for refinement.
CASP organizers
|
|
| Human/server targets are over | (2008-7-12) |
|
We have finished releasing human/server targets for prediction in CASP8. Last week we released more targets of this type than usual in order to give you full 3 weeks to predict each of them. All in all, we released 57 human/server targets in 11 weeks of the experiment or roughly 1 target per business day. Next week is the last week of CASP8 target release and we will be releasing "server only" targets.
|
|
| T0472 - canceled for human predictions | (2008-7-9) |
|
Target T0472 has been canceled for human predictions as the corresponding structure had been released by the PDB. The target is turned over to the "server only" status and will not be assessed in the human-track part of CASP8.
|
|
| Refinement targets - 3 proteins (4 domains) | (2008-7-8) |
|
Please, pay attention that 3 targets for the refinement experiment have been released today. The details of the suggested refinement targets are provided in the pdb files for the starting models and in the "Additional information" field for the targets.
|
|
| T0465 - switching to human/server target | (2008-6-30) |
|
We are switching T0465 over to human/server target. In order to provide enough time for human-regime modeling, we are adding 10 extra days to the prediction window for this target so that it will expire for human prediction on July 21.
|
|
| No targets on July 4 | (2008-6-27) |
|
Next week we will have no targets released on Friday. Next target release date after the holiday - July 7.
|
|
| Extended timeframe groups | (2008-6-24) |
We are launching an additional service designed to help test methods requiring more time than the usual 3 weeks allowable in CASP8. Registering an additional “extended timeframe” group will provide a time stamp certifying the PRE-diction nature of the submitted models as well as CASP numerical evaluation results in the familiar format. Please note that the predictions sent by an "extended timeframe" group WILL NOT be part of the CASP8 experiment, WILL NOT be included in the final results tables and WILL NOT be assessed by the CASP assessors.
Please, use the appropriate registration form ONLY if you feel that you will need more time than the usual 3 weeks for submitting predictions. Predictions from the "extended timeframe" groups will be accepted for 6 weeks total. If the experimental structure becomes publicly available during the extended time window, the predictions accepted after that date will be discarded. Prediction Center does not fully control the extended deadlines and there is real risk that already submitted predictions will be discarded.
|
|
| CASP8 update. New fold targets appeal. Refinement targets. | (2008-6-10) |
General participation statistics
We are in the middle of the CASP8 prediction season. Interest in the CASP experiments remains high. We have 221 groups including 121 servers actively participating in CASP8. More than 28 thousand predictions have been collected by the Prediction Center in the first 5 weeks of the experiment.
Targets
Fifty nine targets (including 28 human/server targets) have been released during the first part of the CASP8 experiment. As you may have noticed, there have so far been few small ‘new fold’ or very remote fold proteins suitable for testing non-template based modeling methods. This is starting to look like a problem, since, obviously, it's very important the community does find out how those methods are progressing.
So we are writing with an urgent appeal to help find target or targets of this type (typically 150 residues or less). Larger proteins with suitable semi-autonomous domains would also fill this need. For those, we need the relevant domain boundary information to be provided with the target. Please go and pester your friendly experimental colleagues, and see what you can come up with.
Refinement targets
The first target has been released for model refinement (TR389). The experimentalists kindly provided us with the coordinates for this target ahead of time. After the automatic evaluation of the models we want to suggest you one of the best models for refinement. Please, pay attention that the refinement target names start with "TR" instead of "T0" for regular targets.
CASP organizers
|
|
| FORCASP migration | (2008-5-19) |
We are moving FORCASP discussion forum from the forcasp.org website to the predictioncenter.org site (link FORCASP forum is added to the menu). Old posts remain available at the forcasp.org website.
CASP organizers
|
|
| Prediction in non-tertiary structure categories | (2008-5-16) |
Some targets in CASP8 Target List are marked as 'Human/server' and the others as 'Server only'. We have informed you that the human-expert groups should concentrate on predicting 3D structures for the Human/Server targets as the assessors will be comparing human-expert groups on this subset of targets. We want to clarify that this requirement pertains to the tertiary structure predictions only. Human-expert groups are expected to submit predictions for all targets (both "Human/Server" and "Server only") for the categories other than TS.
CASP8 organizers
|
|
| Request for CASP8 targets | (2008-5-7) |
Dear CASP8 participants,
The CASP8 prediction experiment has just begun, and as always, it will only work if we have the help of the experimental community in providing targets.
Many targets have been provided by structural genomics centers in the last two CASPs, and expect to that will be the case again this time. However, we still very much need targets from the broader structural community. We need all sorts of targets but in particular, there are several sorts of targets we expect to be short of.
1. We expect the number of ‘new fold’ provided by the SG centers will be smaller this time, and we really really need these to test ‘ab iniito’ target prediction methods. So if you have one, we would be very very grateful if we could use it as target.
2. We want to improve our test set for assessing disorder prediction methods, so if you have structure that contains significant disorder, we would very much appreciate using it as a target.
3. In all the CASPs to date we have only ever had one membrane structure target, so if you have you could make available, that would be great.
CASP only succeeds because of the support and co-operation of the members of this community, and we very much need your help.
The time table is similar to other CASPs: The prediction season opened at the beginning of May, and will run until the end of July. We will release targets continuously throughout that period, as evenly spaced as possible, and hope to have about 100 targets altogether. Each target will be available for prediction for a period of three weeks, although in some cases we request a longer period to allow it to be used to test refinement methods. It is of course important that there not be any kind of public release of the experimental structure (including things like pictures on web pages or abstracts) until after the predictions for that target are closed.
It’s fairly simple, with just two things to bear in mind. First, because of the timing framework, there needs to be at least a month between the submission date and any release of the structure. Second, we ideally need the experimental co-ordinates by the beginning of August and definitely by the beginning of August, so that the predictions can be assessed. At that point, they can be kept confidential if necessary, though we would like to provide them to predictors of your structure at the beginning of November at the latest, so that can see how well they have done. Participants would also usually like to be able to show slides and discuss their predictions at the meeting at the beginning of December.
So, if you have any thing suitable, we would be most grateful if you would enter the data on the web form: http://www.predictioncenter.org/casp8/targets_submission.cgi
OK, that's about it. Please get in touch if you have any questions or concerns. Hoping for lots of targets....
CASP8 organizers
|
|
| CASP8 started | (2008-5-5) |
We have released the first two CASP8 targets. Good luck everyone in predicting these and all the subsequent CASP8 targets.
Please, pay special attention to the column "Type" in the Target List. In the current CASP experiment we are planning on releasing between 50 and 60 targets to be evaluated in the human and server track and many more targets (depending on target availability) to be evaluated in the server-only track. Assessors will be comparing all participating predictor groups on the subset of Human/Server targets and, additionally, the server groups on all released targets. The human-expert groups should concentrate on predicting structures for the Human/Server targets. Those, wishing to take the challenge and predict all of the released targets (including the "server-only" ones) are welcome to do so but the evaluation emphasis will be placed on the "Human/Server" targets. Subject to availability, we will give priority to releasing targets containing low homology domains in the Human/Server track.
Server predictions for all targets will be made publicly available immediately after acceptance by the CASP verification system. Receiving status of server predictions can be checked from the main CASP8 web page by following the link "Server Predictions." Please note that tarballs of all 3D models (with AL predictions already converted into coordinates) will be posted at the same place usually with one-day later. The electronic address of these tarballs will be sent to the quality assessment (QA) servers 4 days after the original target release. QA servers will have an extra 3 days to submit QA predictions.
|
|
| CASP8 server dry run - April 28 | (2008-4-28) |
On April 28 we will have the last round of dry runs for the servers. Please get your servers ready.
The first CASP8 targets will be released on May 5.
|
|
| CASP8 server rules | (2008-4-2) |
CHANGES relative to CASP7
There will be just minor changes to the rules for server participation in CASP.
According to the suggestions of CASP7 predictor meeting, in CASP8 we are extending the window for accepting server predictions. Servers will now have full three days to return the prediction. No additional time for corrections will be allotted but corrections will be accepted within the original 72 hour window.
Also, according to another recommendation of the CASP7 predictor meeting, we will have two different sets of targets for evaluation in the human-and-server track and in the server-only track. We plan to release as many targets as possible for evaluation in the server-only track and a subset of around 50 targets from among these for the human-and-server track.
For the servers participating in the model quality assessment category, we will be sending a link to the tarball of accepted server structure predictions on a particular target in 1 business day after closing the window for server 3D predictions. The quality assessment servers will have another 3 days to generate the model quality assessment prediction.
DRY RUN
We will be testing connectivity with registered CASP8 servers starting on April 14, 2008. To test your server, we will send you an old CASP7 target (T0283). Please respond as you would for CASP8 targets. Of course, these predictions will not be part of the CASP8 experiment.
If your server participated in CASP7, we expect the test to be just a formal recheck. For new servers: we will be sending 3 variables to your server's submission URL: the SEQUENCE, the TARGET-NAME and the REPLY-E-MAIL (where to return the results). For the quality assessment servers we will be sending the TARBALL-LOCATION variable instead of (or in addition to, if you specify so) the SEQUENCE. Please, make sure you provide the required parameters in your CASP8 server registration.
IMPORTANT THINGS TO REMEMBER
1. Target distribution procedure for servers.
You will be receiving CASP8 queries from the distribution server in Davis and the results will be collected directly by the Prediction Center. At the end of each week we will notify you about how many targets we plan to release the following week. Targets are planned to be released on business days only. We plan not to exceed a load of 3 targets per day for servers. Targets will be made available and requests to all servers will be sent around 9am PST. The submission engine will resend the query if it encounters obvious connecting problems (network timeouts, 'no response' etc.). Failures that go beyond that require special attention, but we'll make every effort to notify server curators ASAP if we suspect something is not working. The facility that allows checking data flow status by server predictors will also be available.
2. Server submission format.
Please submit your predictions in the standard CASP format. Your submissions will be immediately verified at time of submission. CASP warning and error messages will return directly and immediately to you, and will refer to the material you submitted, allowing you to respond in close to a server time frame, if necessary. Please note that, as in previous experiments, no prediction will be finally accepted by CASP until it passes CASP format verification.
Below we summarize differences in format of server submissions from that of human groups.
2.1 Header AUTHOR should contain your group name instead of the 12-digit predictor code, e.g.
AUTHOR 3D-JIGSAW
or
REMARK AUTHOR 3D-JIGSAW
2.2. Header SCORE in your prediction is allowable and optional (a real number reflecting self-estimation of the prediction quality).
2.3. For servers, it is allowed to have several models for the same target in one file. If your server returns only one file with several models, we will preprocess such an entry. You may add only one set of required header fields (PFRMAT, TARGET and AUTHOR) at the beginning of the file. Our script will split a multiple model file into separate files (one model per file) and only then we will send each model separately to our verification software (models numbered 6 and higher will be ignored). Error messages (if any) will be sent for each model separately. So you will be able to correct models (if needed) separately.
3. The timeframe for server prediction acceptance is the following. All server groups have 72 hours since target distribution to return the prediction. Error notifications will be sent to the server contact person's email address. If your prediction doesn't pass the CASP verification system, you will need to correct your format manually and resubmit the rejected prediction within the original 72-hour window. To do this, you will need to include the following line in your revised prediction:
REMARK Corrected server submission
and send it by email to: servers@predictioncenter.org .
After 72 hours no server predictions will be accepted or altered in any way. If you expect your server to work on prediction generation more than 72 hours, then we would advise you to register a regular (not server) prediction group for CASP8 instead.
You will be able to check the submission status of your prediction through the Server Submission Status Page or through the Model Viewer facility.
|
|
| CASP8 registration | (2008-3-22) |
The registration for CASP8 experiment is now open.
To register for the experiment, you will need to register with the Prediction Center (PC) first (if you haven't done this yet). At this step we will collect the basic registration information to be used for your participation in CASP8, all subsequent CASP experiments, between-CASP initiatives, and local services. The information provided by you will not be shared with anyone outside of the CASP system.
After you register with PC and login to our website, you will be able to register for the CASP8 experiment (a separate registration form under the CASP8 menu). You will find five separate registration forms there. Please, read the registration instructions and use the appropriate form.
We ask that you register with the Prediction Center at your earliest opportunity as we plan to release future CASP8-related news mainly through our website. We recommend that you select an option "Subscribe for Prediction Center newsletters" as in this case you will be receiving emails every time the news are posted on our website. This functionality was just recently added to our registration form. So, if you have registered earlier, please go to "My personal Data" link and select this option.
Also, we ask server group leaders to register as soon as possible as we are planning on testing the connectivity with the registered CASP8 servers starting April 14, 2008. Server participation rules and the details of the dry run will be posted in the Message Board section of our website shortly (this section is available from both the Prediction Center home page and the CASP8 home page).
|
|
| CASP8 preliminary timetable | (2008-2-1) |
Registration for the experiment will start in the last week of March 2008.
Prediction season will run from early May through early August 2008. The first prediction targets will be released not earlier than May 5, 2008; the last prediction targets will be released not later than July 18, 2008; prediction season will end not later than August 1, 2008.
According to the recommendations of the CASP7 predictors meeting, we will be releasing about 50 targets for evaluation in the human and server track and as many targets as we can for evaluation in the server-only track (including "human and server" targets). The human-expert groups wishing to take the challenge and predict all of the released targets (including the "server-only" ones) are welcome to do so but the evaluation accent will be placed on the selected targets. Subject to the availability, we will give priority to targets containing low homology domains for inclusion in the human-expert modeling experiment.
The meeting to discuss results of the experiment will be held on the island of Sardinia, Italy, December 3-7, 2008. The meeting will take place at Hotel Setar , near Cagliari. The hotel is located in Quartu S. Elena, 10 kilometers from Cagliari and 15 from the Cagliari airport.
|
|
| CASP 7.5 meeting | (2008-2-1) |
|
Together with CNIO, Madrid, we are organizing the next between-CASP meeting (CASP7.5). It is going to take place in Madrid, Spain in April 2008. The idea of the between CASP meetings is to update the progress of structure prediction as measured by the CASP experiments. In contrast to the full CASP meetings that are exclusively for participants and developers, this meeting will be open to all interested scientists. All interested please check the meeting web page: http://ubio.bioinfo.cnio.es/casp/index.html
|
|
