I said the computational attributes of the protein are intrinsic to the biological system already.

There's a discussion about proteing folding here, that anyone in this project needs to read, and thoroughly understand:

Protein Folding and organic computation:

http://www.iidb.org/vbb/showthread.php?t=107159&page=2&pp=25

 

hi there!

I' ve just joined the betterhumans - rosetta - team. don' t worry, i' ll never ever leave you guys alone, but i have to ask just one question: what about simap? that project approaches the protein folding issue from another angle (it' s neither better nor worse, just different) than rosetta@home. wouldn' t it be good if we had a betterhumans team that covers both projects? right now my boinc settings are 50 - 50.

simap details: http://en.wikipedia.org/wiki/SIMAP

rosetta@home details: http://en.wikipedia.org/wiki/Rosetta%40home

Hi Ketzl, good to have someone from the (almost) neighborhood in the team (Grüße nach Österreich ;-). Despite being located _very_ close to where simap is headquartered, ;-) personally I'll stick with 100% Rosetta, mostly because I think I understand enough of what they are doing to be convinced that this is important work which eventually we all may benefit from - and also because I like the direct interaction with the team (it almost feels like family by now ;-).

I am not sure you already found this thread in the Rosetta forum which discusses possible synergies/cooperation between Rosetta and simap, with posts by Rosetta staff and I think someone associated with simap. So I certainly understand your decision to split between the two projects - and if you or someone else feels like creating a betterhumans.com team over there, I'd say go for it - though perhaps you should wait till you are sure you won't be the only team member of the new team. I'd even be willing to change the title of this thread to also include simap in the title. ;-)

Chironian:
I am not sure I completely understand the point you are trying to make. Well yes, one could say that a protein computes itself. That would be the view that the universe consists of computation - or of Mathematics if you will (mathematical Platonism; I am somewhat inclined to think along those lines). If we had a sufficiently powerful quantum computer we might even be able to compute the folded shape of a protein within a time that is not very much longer than it takes the protein to fold in nature - and our current efforts would look somewhat pathetic at that stage. But such quantum computers don't currently exist and it is not even certain that they ever will. So it seems to me we shouldn't just 'wait' till that future time (which may never come) but do our best with the currently available hardware.

Pace:
If one defines "solving the protein folding problem" as "computing the folded shape of a protein from its sequence of amino acids to sufficient accuracy to be biologically useful" then I believe Rosetta indeed has a chance to eventually solve that problem (I made the point in the Rosetta forum that that might be worthy of a Nobel prize ;-).

I am not sure those many physicists doing n-body simulations would agree with your statement that the many body problem is intractable, ;-) and I am also not sure it is a very good analogy for Rosetta. Given enough computing power it would certainly be possible to find the lowest energy conformation of a protein. The time needed to find the exact minimum will grow exponentially with the size of the protein, though (it is hard to beat an exponential), but it is hoped that sufficiently accurate approximate solutions can perhaps be found in less time. The usual text book example for this type of "np-complete" problems is the "travelling salesman problem" (find the shortest itinerary connecting n destinations) and there are algorithms (simulated annealing, genetic algorithms...) that find useful approximate solutions (a sufficiently short itinerary) in less than exponential time. Actually, simulated annealing has been tried by Rosetta and I think they once had someone trying to implement a genetic algorithm approach in Rosetta. Coming back to the many body problem, you were of course referring to the fact that the movement is chaotic, thus requiring infinitely accurate initial conditions and computations (both of which is impossible), so this is different from why what Rosetta is doing is so hard...

Aidan:
Thanks for the additional info and actually the 'business' I work at is a scientific research institute - but one working in an area very much removed from what Rosetta is doing (one might say lightyears removed ;-) and I suspect that you overestimate the scientific literacy of the general public (which is who we want to get interested in crunching for Rosetta)... ;-)

Hoelder1in:

...I suspect that you overestimate the scientific literacy of the general public...

 

I hope that isn't true...though, sadly, I have seen enough examples to admit it may be. Still, I spent most of my time in school sleeping, dreaming up ways to con people out of their money, or arguing with the principal about why I didn't show up at all last Thursday.  And I still picked up a few things.  I'd hate to think there were people who payed even less attention than I did...much less that they may be the majority!

Hoelder1in:

...the 'business' I work at is a scientific research institute - but one working in an area very much removed from what Rosetta is doing (one might say lightyears removed ;-)...

I wish I was clever enough to offer a good guess, but alas I cannot decode the witticism...

 

BTW: Welcome to the thread and the team Ketzl.

Hi everybody. I just joined the team.

Some other BOINC projects that seem transhumanist-relevant are climateprediction.net (global warming), QMC@Home (improving quantum chemical simulation techniques), Spinhenge@Home (studying nanoscale magnetism), and uFluids@Home (studying microscale fluid behavior, with applications including better rockets).

Yep, I'm sharing my CPU cycles between Rosetta and ClimatePrediction.

MikeGR:

Hi everybody. I just joined the team.

WOW, this is a first:  two new team members on a single weekend. Welcome Mike - you seem to be one of those who were around since the early days of Rosetta (they call that 'class of 2005' in the Rosetta Cafe ;-).

Hoelder1in:

MikeGR:

Hi everybody. I just joined the team.

WOW, this is a first:  two new team members on a single weekend. Welcome Mike - you seem to be one of those who was around since the early days of Rosetta (they call that 'class of 2005' in the Rosetta Cafe ;-).

 

I didn't know there was such a "class". Cool.

 
Years ago I was doing some seti@home. When other distributed projects started surfacing, I moved to cryptography, and then science. I did some folding@home, lots of predictor@home, lots of climateprediction. I've been with Rosetta@home for a while, though I wasn't always faithful; sometimes I dropped it in favor of other projects... But now that I understand it more, I think I'm going to stick with it.


 

Hoelder1in,

I stand corrected. What you said was what I was trying to say.

When I said intractable what I should have said was, "--hard to solve (Or perhaps impossible to solve perfectly) but not impossible to approximate or solve in special cases." The N-Body problem is not impossible to deal with either. Approximations work well enough and in many cases that's all we need. Most of the physics in simulations of galaxy formation (which is a case of the N-Body problem, relativistic or Newtonian. ) are approximations but they are very informative ones.

The point I was badly trying to make is that we don't have to understand nonlinear systems completely and perfectly in order to get very good at weather prediction, climatology, drug refinement, artificial intelligence, protein folding, air foil design, etc etc. etc. I was just trying to head off Chironian's vague attempts to invalidate the premises of Rosetta (And by implication, climatology, astrophysics, etc. etc. etc. His earlier statements suggested to me that he doesn't have a good understanding of nonlinear dynamics and I was trying to clarify that.). 

I think you and I are on the same page on this even if I'm not so articulate about it.

I've been juggling computers around a bit this past week. Converting away from buggy 64-bit windows, setting up and taking down firewalls, and physically relocating a few machines. I've been getting quite a few errors out of R@H. The compute errors I can trace back to multiple restarts (installing software) and the one machine with multiple users and a shared installation--which I have now converted to a service install to avoid killing Rosetta repeatedly. But the "validate errors" have me stumped. Is there something I could have messed up while reinstalling boinc? Or another system stability issue which might cause such errors? The machine in question is now in Philadelphia, so I can't just pop over to have a look.

I have to admit that my understanding of the mechanics of BOINC is limited. My next stop is a long reading session in the R@H forums, but I thought I'd run this by the big brains here first.

AidanSonoda:

But the "validate errors" have me stumped.

Me too. ;-) All I can say is that others seem to have had the same error (a Google search for "Rosetta score is stuck or going too long" threw up three results - two of those on Ralph). I didn't see any suggestions how to fix the problem, though, and I have no idea why it just happens on one of your computers. Perhaps someone on the Rosetta forum could help (Chu from the Rosetta team seems to post a lot in the error reporting thread) if you describe the problem over there once more...

Well, the errors seem to have spread to all but one of my machines, including ones I didn't meddle with.  No useful answer at the forums, though the timing suggests it is probably related to the update of the Rosetta program (to 5.34).  I posted about it under "Problems with 5.34" but no useful responses yet.  Am getting other errors as well, but not consistently and now across many machines.  Hopefully it is a problem on their end, or someone will be able to suggest some change I need to make at mine. (other than hardware there is essentially nothing unique about the one host of mine that remains error free). Thanks for taking the time to help. 

Well, it certainly seems like the team is stuck at 15 members.

 

I have a friend who might be interested and I'll try to get him to run Rosetta. Anybody else has recruiting opportunities? Even if they don't end up joining the team, just running Rosetta is very important. 

I'm still trying to get some spare cycles from two of the linux clusters I've recently set up. The owners are running SETI 100%...which aggravates me, since there is much more important (and valid) science to be done. I think they like the prettier graphics. I'll keep pestering them though. Otherwise I think it is essential to try to recruit from forums like this one. I agree with Hoedler1in, that this is an untapped resource for this kind of research. Personally I'm not a forum kinda guy...this is pretty much the only one I pay any attention to...and I came here from the R@H website simply because I was looking for a team to join. Perhaps some of the more web savy members could do some polite proselytizing for us.