Posts by buren

1) Message boards : Number crunching : Running in "high priority" with no real reason (Message 59164) Posted 29 Jan 2009 by buren Post: I noticed that Rosetta often is switched to high priority at the beginning of the processing of any workunit. But as you can see 18% of that unit only tool 4,5h. So all in all it should roughly need another 20h of work. With the deadline being 9 days ahead that would only be 2h a day and BOINC is run much longer. After a day or so things get back to normal and the high priority is disabled. I'm using Boinc 6.4.5 and the last benchmark turned out about 2500 (Whet) and 4200 (Dhry). Is that a problem with Rosetta or is it Boinc?
2) Message boards : Number crunching : Problems with minirosetta version 1.+ (Message 51313) Posted 11 Feb 2008 by buren Post: Yes, it's taking more RAM. I'm running one of each right now. Peak for 5.93 is 114,764K and for mini is 185,424. Current usage is 103M for 5.93 and 174M for mini. Same here, 167MB (172 max) for Mini vs 154MB (155 max) for standard. But Mini only takes 166MB of virtual memory, while the standard Rosetta takes 230MB. I don't know if you can just add up those two numbers to get the total memory usage of 334MB vs 484MB. But this probably means anyway that Mini actually takes less RAM. first one came in 5 credits under claimed, the second one came in at 8 under claimed. kinda hurting my average with these. I almost always get a little less credit than claimed. 8 credits out of 350 is only about 2% anway, so it shouldn't matter much.
3) Message boards : Rosetta@home Science : Plotted results not working (Message 50338) Posted 5 Jan 2008 by buren Post: In the Q&A it was said the plotted results often fail. I'm mainly interested not in the plot, but simply in checking what lowest energy/rmsd was calculated during the run. Is there any way to get this data from the files on my own HDD? In that case I could go w/o the plotted results a while longer... :)
4) Questions and Answers : Web site : Result plots not working? (Message 50069) Posted 26 Dec 2007 by buren Post: Any news regarding this matter? I really miss the possibility to check how good the predictions were.
5) Message boards : Rosetta@home Science : Plotted results not working (Message 49939) Posted 22 Dec 2007 by buren Post: Ok, I didn't check the other forum after posting here. So it seems it's in the work. :)
6) Message boards : Rosetta@home Science : Plotted results not working (Message 49830) Posted 20 Dec 2007 by buren Post: Am I the only one that's experiencing this problem?
7) Message boards : Rosetta@home Science : Plotted results not working (Message 49829) Posted 20 Dec 2007 by buren Post: Am I the only one that's experiencing this problem?
8) Message boards : Rosetta@home Science : Plotted results not working (Message 49696) Posted 14 Dec 2007 by buren Post: I posted that in the website q&a but there isn't much traffic, so I'll post it here again. Hope that's okay. :) Since some weeks I cannot access the plotted results site. It gives this error after some time: "Sorry, the data requested does not exist." Is that a known issue and do you have any idea when it'll be fixed? Especially being able to track how good an rmsd was found is interesting. Maybe the client itself could show that value?
9) Questions and Answers : Web site : Result plots not working? (Message 49638) Posted 12 Dec 2007 by buren Post: When trying to visit the Results page I always get a time-out. The result page opened from within the client works, but you can't view the plotted results and so far found minimum RMSD there. Is that a known problem?
10) Message boards : Rosetta@home Science : Are really low energy structures worth it if the RMSD isn't so good? (Message 49039) Posted 25 Nov 2007 by buren Post: Okay, with DC it probably really doesn't matter how many models a single computer runs for each structure/WU. I forgot that other computers can resume the same WU with different models. So the total number of tested models per WU stays the same. Because if one computer only tests 5 models per WU another computer might check the remaining 45. Does anyone know how much models per structure are actually tested? Or does it depend on the structure?
11) Message boards : Rosetta@home Science : evolutionary algorithm (Message 49038) Posted 25 Nov 2007 by buren Post: Okay, I thought it would mean the way to predict the correct structure would follow evolutionary strategies. But I guess you're right, it often means the code itself evolves. But somwhow what we do currently is kind of that. There are many different models running on all the participants computers and only those models that come close to the real structure are taken into account while the others are discarded. Something like a manual evolution.
12) Message boards : Rosetta@home Science : Are really low energy structures worth it if the RMSD isn't so good? (Message 48953) Posted 22 Nov 2007 by buren Post: Okay, I read about being able to set the time per WU but couldn't find the preferences and so I thought it was a fixed time. What's prefered, more WU less detailed or less WU more detailed? I guess the longer the WU runs the better you know if the model works well, but probably there is a time after which you mostly know if the model is useful or not. So setting the time too high might waste resources. Most of my models did end with RMSD of around 10 after 2h, I don't know how exactly the RMSD is calculated and what RMSD are acceptable but at least the structures still looks way different to the eye from the real life structure. So have there been any improvements so far and how good must a model be i.e. how close to the real structure to be considered "it".
13) Message boards : Rosetta@home Science : evolutionary algorithm (Message 48911) Posted 21 Nov 2007 by buren Post: Isn't the way how the folding works currently sort of evolving? The model doesn't try all possible structures as that would take too much time even with DC. But it checks different ways to continue from a start structures and works its way from there. That's probably the main reason why why currently only do test runs for the model. If the model would test all structures then it would 100% find the lowest energy one finally. But since it doesn't test all but chooses based on some kind of evolutionary model we (well not exactly we) have to make sure that even though not all structures are tested, still the right one is found. So I think such a model is already in use. But still somewhat in beta testing.
14) Message boards : Rosetta@home Science : Suggestions for improvement thread (Message 48872) Posted 20 Nov 2007 by buren Post: I posted this in an own thread already, because I didn't know such an improvemtn thread did exist. So here's my suggestion again: If I did uderstand the experiment right, currently what we do is not actually discovering the structure of a proteine, but checking if the model is capable of doing so. So it'd be interesting to be able to see if the model currently running got close to the real data by displaying the known lowest energy. Maybe even the percentage already reached. Also it'd be nice if in the RMSD window the zero line was marked. So one could easily how close to the real one the model works. And btw, does anyone know how the project is doing? Any chance of getting to predict real structures anytime soon? I read the journal but there's lots of entries and hard to find what I was looking for.
15) Message boards : Rosetta@home Science : Are really low energy structures worth it if the RMSD isn't so good? (Message 48865) Posted 20 Nov 2007 by buren Post: Currently the models finish quite fast on modern PCs, about 1-2h each. Are the times only that short for the test runs or will they be that short with real untested structures as well? In that case it won't take that long to test a whole lot of proteines, making the project not really dependened on DC, so I guess it will take longer with the real ones. But why are the test runs that short?
16) Message boards : Rosetta@home Science : German questions (Message 48864) Posted 20 Nov 2007 by buren Post: Though english ist not the first languange for most people, in western civilations (with that I mean Japan and the such as well) and thus the majority of the internet the one language most people can speak I think is english. So if you want a text to be understood by as many people as possible, choose english. For example, the most common native language in the EU is german. But still english give you the best results in holiday or whereever since it's the language most people as least know a little. What I want to say is, while it certainly is nice of there are translations, english still is the lingua franca of the internet no matter if the number of actual native speakers is relatively low.
17) Message boards : Rosetta@home Science : Why are there so many turquoise points? (Message 48863) Posted 20 Nov 2007 by buren Post: It seems the lowest RMSD is at the very left. So the optimal result (RMSD=0 and energe=min) would be in the lower left.
18) Message boards : Rosetta@home Science : Show the empirically found lowest energy in client (Message 48814) Posted 19 Nov 2007 by buren Post: If I did uderstand the experiment right, currently what we do is not actually discovering the structure of a proteine, but checking if the model is capable of doing so. So it'd be interesting to be able to see if the model currently running got close to the real data by displaying the known lowest energy. Maybe even the percentage already reached. Also it'd be nice if in the RMSD window the zero line was marked. So one could easily how close to the real one the model works.
19) Message boards : Rosetta@home Science : Why are there so many turquoise points? (Message 48799) Posted 19 Nov 2007 by buren Post: Thanks, I was confused because it said "The turquoise points are from calculations we did here where we cheated and started with the actual structure." on this page: http://boinc.bakerlab.org/rosetta/rah_top_predictions.php And where is the "zero line" (don't know the correct english term) for the RMSD in that picture? Lowest energy is at the very bottom, but RMSD should be zero for the best result. Is it on the very right?
20) Message boards : Rosetta@home Science : Why are there so many turquoise points? (Message 48787) Posted 18 Nov 2007 by buren Post: Okay, but each model refers to the same protein, or doesn't it? So when the dots show the "real" data there should only be one. Or how are those green ones acquired? It also looks like the the model running on my computer adds green dots all the time. Is that true? But I think I read the dots are from earlier experiments.

Next 20