Posts by eberndl

1) Message boards : Number crunching : Problems with version 5.96 (Message 53827) Posted 19 Jun 2008 by eberndl Post: Hello, I've had at least 3 units go to 100% and then just sit there. I'm running the linux version of 5.10.45 on Ubuntu 8.04. The most recent unit I've had to kill is t405__CASP8_JUMPAB_TYPE2_RES81to192_SAVE_ALL_OUT_BARCODE__3785_69982 (http://boinc.bakerlab.org/rosetta/result.php?resultid=172177470) My computer is a dual core with 2GB of RAM...
2) Message boards : Number crunching : optimized apps. (Message 20264) Posted 15 Jul 2006 by eberndl Post: Sorry, I guess I meant "no unofficial optimized apps"
3) Message boards : Number crunching : optimized apps. (Message 20262) Posted 15 Jul 2006 by eberndl Post: It has been discussed before and the official answer from Dr. Baker is NO
4) Message boards : Number crunching : Stuck at 74.101% using 0.000% CPU (Message 18930) Posted 19 Jun 2006 by eberndl Post: Carlos, If Rosetta wasn't using your CPU time, what was?? Sometimes another program gets stuck and steals all the cycles. Another thing that has happened to me once or twice is that the system idle process takes the cycles, and BOINC makes no move to use them. When this happens, I shut down BOINC, and then restart it (ie give it a kick in the pants)... that almost always gets it going again. Good luck!
5) Message boards : Rosetta@home Science : T0300 dimer - possible? (Message 18790) Posted 16 Jun 2006 by eberndl Post: I would assume that the hope is that Rosetta can eventually, because from what I understand, Rosetta is for "Protein folding, design and docking", where docking means protein-protein interactions. It is likely that the conformation of the dimer would be slightly different than that of the monomer, especially if the protein forms a heterodimer, that is, Protein A dimerizes with a different type of protein, B. To understand the full functionality of the protein, we'd need to know what the dimer (or tetramer, or octomer...) looks like, but even in the monomer form, certain motifs (like well known binding domains or cleavage sites) may become visible, helping to prove the function or mechanism of action of the protein.
6) Message boards : Number crunching : Have you ever ran out of quequed work units? (Message 17088) Posted 25 May 2006 by eberndl Post: I have, but only when I've done too much Rosetta and I have to catch up on LHC... I don't think that the Rosetta project has EVER run out of WUs, this is partially because even if they re-send WUs, they still get a different random seed once they reach the individual computer. Right now, Rosetta produced about 31TFlops, if I remember correctly, they want to be in the 100-150TFlop range. Until then, we will NEVER have to worry about running out of Rosetta units =-) (unless the splitter breaks down or something like that)
7) Message boards : Rosetta@home Science : Just a question. (Message 16419) Posted 17 May 2006 by eberndl Post: Chuck, What we're doing now is very useful and does lead to major breakthroughs. Up until now, we have been running Rosetta against proteins whose structures were already known. Someone else had already figured them out, and we were doing this because we were basically calibrating and improving the Rosetta search technique. The units we're doing until mid August (or so) are for a Bi-annual competition called CASP, and we're trying to predict structures that have not yet been published. That means that we don't know what they're supposed to look like, but various (traditional) labs are nearly done determining the structures. We have to return our computed structures before they get published to see how good Rosetta's (or Protein Predictor's, or any of the other groups we're competing against) system really is. The purpose of Rosetta is to one day be able to feed in the sequence and quickly and accurately get the structure back. This is a test of that system. This is how Rosetta can show how good its prediction system is.
8) Message boards : Rosetta@home Science : Comments/questions on Rosetta@home journal (Message 16262) Posted 14 May 2006 by eberndl Post: Today (Friday) we have closed accepting server predictions for the first target, T0283. What should be done with T0283 WUs that are still on our machines? I'm with Robert on this one... should we be aborting the unfinished 283s??
9) Message boards : Rosetta@home Science : Comments/questions on Rosetta@home journal (Message 15593) Posted 5 May 2006 by eberndl Post: So, even if you had the BILLION proteins all crystilized, it would STILL take WAY too long (billions of hours) to X-Ray all of them. Feet1st, you're right it does take hours to Xray each crystal, but it can take weeks or MONTHS to figure out how to crystallize a protein for the first time. And once you have the xrays... you have to run that through a massive computer program (another couple hours), and what that will give is an electron density map, which looks like this: But all those solid yellow lines have to be added BY HAND to the density map. Once this first estimate of the protein is obtained, they do a theoretical Xray on their model protein, and look at the differences between the actual (from the crystal) and experimental (from the model) scatter graphs, and then they go through and try to reduce the differences as best they can. again, they have to do this by hand. As for chemical reactions... I believe that is how they know the sequence that the amino acids comprising the protein are in. But, ya, perhaps they could do some other chemistry on it to infer some broader info. to get clues as to how they are arranged. There are 2 main ways to figure out the sequence of a protein: Get the cDNA and convert it to protein, or take the protein and do tandem Mass Spectroscopy on it, which can give you the actual protein sequence (except that they can't tell leucine and isoleucine apart due to their identical sizes).
10) Message boards : Rosetta@home Science : Comments/questions on Rosetta@home journal (Message 15532) Posted 4 May 2006 by eberndl Post: Hugo, From what I understand, the whole point is for Rosetta to be able to predict the structure WITHOUT using the X-ray data. As it stands, that data (or data from protein NMR) is what provides the images in the "Native" box of the screen saver. Although being able to start with a good approximation is very important, it's also important to be able to start from scratch, because most proteins don't have homologues that have been solved yet.
11) Message boards : Rosetta@home Science : Where are amino acids built? (Message 15471) Posted 4 May 2006 by eberndl Post: The body can build 12 of the 20 amino acids inside the cell... this does require enzymes which are made of proteins though. Th remaining 8 amino acids are essential, meaning that we don't have enzymes to build them, and so we have to get them from the food that we eat. Protein rich foods (like meat, beats, rice, wheat and tofu) get broken down into individual amino acids which are then transported in the blood to all the cells of the body. Any amino acids that are not needed get broken down for energy, or stored as fat (we cannot store excess amino acids). Amino acids are also recycled within the cell. Some proteins are only needed when a cell is dividing, for example. Once the cell is done dividing, that protein gets broken down (by other proteins) to individual amino acids. Free amino acids, whether produced by the cell, obtained from food, or recycled within the cell, then hook up with their proper tRNA to build the next protein that you need.
12) Message boards : Rosetta@home Science : Discussion- Proteins and Work Units (Message 15341) Posted 2 May 2006 by eberndl Post: I did some poking about, and here are some sizes from hydrogen to a eukaryotic cell, to give ideas of sizes... amino acids and simple globular proteins are in bold. I'd guess that the globular protein has about 70-110 amino acids in it. # 0.1 nm (nanometer) diameter of a hydrogen atom # 0.8 nm Amino Acid # 2 nm Diameter of a DNA Alpha helix # 4 nm Globular Protein # 6 nm microfilaments # 10 nm thickness cell membranes # 11 nm Ribosome # 25 nm Microtubule # 50 nm Nuclear pore # 100 nm Large Virus # 200 nm Centriole # 200 nm (200 to 500 nm) Lysosomes # 200 nm (200 to 500 nm) Peroxisomes # 1 um (micrometer) # (1 - 10 um) the general sizes for Prokaryotes # 1 um Diameter of human nerve cell process # 2 um E.coli - a bacterium # 3 um Mitochondrion # 5 um length of chloroplast # 6 um (3 - 10 micrometers) the Nucleus # 9 um Human red blood cell # 10 um # (10 - 30 um) Most Eukaryotic animal cells (Source http://ccgb.umn.edu/~mwd/cell_www/cell_intro.html) And I think the smallest protein I've ever heard of is 8 amino acids long (it's involved in labour and delivery so it has to be produced FAST)
13) Message boards : Rosetta@home Science : Concerns: Will someone (i.e. BigPharma) make money out of my CPU time? (Message 14713) Posted 27 Apr 2006 by eberndl Post: Getting on my soap-box for a moment, this is why the big pharms are putting very little effort into things like Beta cell, Islet and pancreas transplants, which are by far and away the best "cures" for diabetes. Instead, they work on better pumps, insulin inhalers and the like. Simply because selling insulin and devices to administer it are a multi-billion dollar per annum cash cow for them that they lose the moment we switch to the transplant cures. It's true that they make a LOT more money from people who are dependant on a product for the rest of their lives. However, if you get a transplant, you're still dependant on drugs... Anti-rejection drugs now. Now, one of the side effects of those anti-rejection drugs is... wait for it... Diabetes. That probably has something to do with why there have been so few islet transplants, though there have been a few (here is the Canadian islet transplant programme and here is some recent research on xenotransplants). The best (ie perfect world) solution would be growing the cells from your own stem cells. Then there's no chance of rejection and no rejection drugs which could re-induce diabetes. My 2 cents.
14) Message boards : Rosetta@home Science : Can proteins aid painkillers (Message 14213) Posted 21 Apr 2006 by eberndl Post: I don't think you could resolve the Cox problems if we knew exactly what it looked like. Think of it this way, you have water flowing down a river that splits into two channels. This upper part of the river is Arachadonic acid. It splits to the thromboxane and prostacyclin streams. But we live on the Prostocyclin, and our house is in danger of being washed away, so we block the Prostocyclin. But then ALL the water goes down the thromboxane, which floods our fields. The arachadonic acid is going to turn into SOMETHING. and if it doesn't turn into something, it will build up, which could be toxic. And you could treat the homophilia with drugs, but hemophilia generally is not considered an "acceptable" side effect to a drug. BUT if we know the protein structures, we might be able to better anticipate side effects, and also make better second and third generation drugs (which generally have fewer side effects). The protein sturctures might also give us new places to try to block something. So, if instead of making a cox 2 inhibitor, we could reduce the prostocyclin by degrading it faster (widening the stream in the example above), you would get a similar effect. knowing the protein that breaks down prostocyclin and making an activator for that protein would be easier if we know that protein's structure. We can't make a magic bullet though, because the system will have to re-balance somehow.
15) Message boards : Rosetta@home Science : Can proteins aid painkillers (Message 14159) Posted 20 Apr 2006 by eberndl Post: Hey Feet1st, Thromboxane and prostacyclin are both derivatives of arachadonic acid (a 20 carbon fatty acid). The links you have are for the the RECEPTORS of these molecules (and the receptors ARE proteins) There's a nice little picture in the Wikipedia here that shows the pathway from arachadonic acid. Basically, COX-2's inhibit the prostocyclin synthetase, so instead of making lets say 50 units of both, you make no prostocyclin and up to 100 thromboxanes. But if you also knocked out the thromboxane receptor (so it can't have any effect on the cell) you'd have reduced chance of heart attack... but increased chance of hemophilia (thromboxane is needed for platelets to form blood clots). Hope that helps a bit.
16) Message boards : Rosetta@home Science : Concerns: Will someone (i.e. BigPharma) make money out of my CPU time? (Message 14158) Posted 20 Apr 2006 by eberndl Post: People!!! We're producing protein structures here, not drug molecules!! When protein structures get published, they go into a common database, that BigPharma DOES have access to. But so do you!! Pharmaceutical companies might chose to target a specific protein that we've crunched (for example prevent it from being functional as a cancer treatment). It's EASIER and CHEAPER if they have the structure, but in all honesty, they can do it without knowing exactly what the protein looks like. Often, they'll look for an end result, then modify the chemical to improve the efficacy, and improve delivery (water solubility is important!!), once they have the drug of interest, they'll figure out the exact mechanism, and it is HERE that the protein structure comes into play. I worked at a small biotech firm that was still working on its first drug after almost 10 years of research (but were in Phase 1 human trials). In this time, they had obviously sold NOTHING.... But it cost about $50 000/day (I might be short a factor of 5) just to keep the place running. They had millions of dollars invested in them from various share holders... this is one of the reasons drugs are so expensive... Because producing effective drugs is NOT easy. Anyway, I HOPE that people can eventually use Rosetta to determine the structures of unknown proteins (note that we already KNOW the structures of all the proteins we're doing... we're doing calibrations and improvements to a system right now). I hope that some of those proteins help produce medicines. If they don't, what's the point?
17) Message boards : Rosetta@home Science : How can I find out what protein my machine is working on? (Message 13679) Posted 13 Apr 2006 by eberndl Post: No problem Larry... Predictor has that site linked up somewhere, so that's how I found it originally. But now I have a question for everyone out there... I now have a unit called 7521_largescale_large_fullatom_relax_dec7521_1_39_1.pdb_437_45_0 How do I find out the protein ID on this sucker?? Any ideas?
18) Message boards : Rosetta@home Science : How can I find out what protein my machine is working on? (Message 13545) Posted 12 Apr 2006 by eberndl Post: Absolutely there is. First you have to go to the BOINC manager and get the protein's identifier... I currently have a unit called ALL_TOPOLOGY_CODES_1vcc_434_295_0 The protein's identifier is 1vcc. Then you can go to the Protein Data Bank and enter in the ID (in the search at the top). You'll then get a page (like this) which has the protein's name, what species it's from, what kind of protein it is (mine's a topoisomerase!!) and which paper published the crystal structure (or NMR structure). Hope that helps!
19) Message boards : Rosetta@home Science : Rosetta video/outreach project (Message 13023) Posted 4 Apr 2006 by eberndl Post: Larelin, It was beautiful... better than some of the animations I've seen elsewhere online. One thing I noticed, though was that during the translation sequence, you can't see the amino acids being added very well. I knew what to expect and had trouble finding them right away. Maybe if the camera angle was on the other side of the ribosome, so you could more clearly see the aa chain growing, it would be a little more understandable?? Of course, this comes from someone whose stick people are barely recognisable...
20) Message boards : Number crunching : no work? (Message 12546) Posted 23 Mar 2006 by eberndl Post: Well... it's BOINC trying to balance the resource share so that over the long run you spend half your time doing each project. BUT if for some reason you couldn't get any SETI units for a little while (for example Wednesday shut weekly maintenance), you'd grab some more Rosetta units so you wtill had something to crunch, and you crunch for say, 5 hours. Then, when SETI comes back up, you won't be able to download Rosetta until SETI has more or less caught up. So for about 5 hours, you wouldn't be able to download Rosetta units. Does this make sense?? Though also quite likely that there are just connection problems as Charles said... Rosetta is apperently running its server to near capacity.

Next 20