Rosetta is by far and away my favorite BOINC project. I think protein folding is very important and the scientists in charge of the project do a good job keeping crunchers, such as myself, in the loop.

One question that has occurred to me is "What have we (the crunchers) accomplished?" In particular, how much more accurate are the predictions from Rosetta compared, to, say Rosetta from 3 years ago? Or, if the CASP targets from 3 years ago were run through the current version of Rosetta, how much more accurate would the predictions be today versus the predictions from the 3-year old CASP competition? I hasten to add that I know so little about this science that these questions might be unanswerable, but to the extent they can be answered I'd be interested in the answers...

---

Rosetta is by far and away my favorite BOINC project. I think protein folding is very important and the scientists in charge of the project do a good job keeping crunchers, such as myself, in the loop.

One question that has occurred to me is "What have we (the crunchers) accomplished?" In particular, how much more accurate are the predictions from Rosetta compared, to, say Rosetta from 3 years ago? Or, if the CASP targets from 3 years ago were run through the current version of Rosetta, how much more accurate would the predictions be today versus the predictions from the 3-year old CASP competition? I hasten to add that I know so little about this science that these questions might be unanswerable, but to the extent they can be answered I'd be interested in the answers...

Hi Mark,

These are good questions of course! we have an in house set of test proteins we use to monitor progress in our methods, and it is clear they have improved considerably in the last three years. The clearest way to see the contributions is through the many scientific papers that we have published over the past several years that were only made possible through the contributions of Rosetta@home participants. We should update the list on the rosetta@home main page; in the meantime a relatively up to date list is available on the "publications" page reached through my group home page (depts.washington.edu/bakerpg).

David

---

Rosetta is by far and away my favorite BOINC project. I think protein folding is very important and the scientists in charge of the project do a good job keeping crunchers, such as myself, in the loop.

One question that has occurred to me is "What have we (the crunchers) accomplished?" In particular, how much more accurate are the predictions from Rosetta compared, to, say Rosetta from 3 years ago? Or, if the CASP targets from 3 years ago were run through the current version of Rosetta, how much more accurate would the predictions be today versus the predictions from the 3-year old CASP competition? I hasten to add that I know so little about this science that these questions might be unanswerable, but to the extent they can be answered I'd be interested in the answers...

Hi Mark,

These are good questions of course! we have an in house set of test proteins we use to monitor progress in our methods, and it is clear they have improved considerably in the last three years. The clearest way to see the contributions is through the many scientific papers that we have published over the past several years that were only made possible through the contributions of Rosetta@home participants. We should update the list on the rosetta@home main page; in the meantime a relatively up to date list is available on the "publications" page reached through my group home page (depts.washington.edu/bakerpg).

David

I pulled up a random pdf file and after skimming thru it didn't see any mention of the Distributed Computing contributions to the paper. Is this on purpose or did I just pick a paper that did not have it in there? I am curious because maybe if the DC contributions were acknowledged in official releases to Scientists perhaps they could see the results possible and maybe more would jump on board and encourage others to do the same. Obviously 'getting the word out' is an important step to getting more and new users over time.

I think one way to do that would be to get more and more Scientists involved and aware causing them to use it in their vocabulary causing their students/whatevers to then do the same. This would then, hopefully, cause other students/whatevers to do the same bring DC into the home from the only people possible, the kids! As kids we all thought we could solve all the problems of the World if we just thought about it enough. As adults we begin to realize the daunting tasks involved. IF we could get the idea of DC being a positive influence and a venue for change maybe we could get it installed on every computer, or at least the Developed World ones! BUT I think if its involvement were acknowledged in official papers, it would start the thought processes flowing.

---

I pulled up a random pdf file and after skimming thru it didn't see any mention of the Distributed Computing contributions to the paper. Is this on purpose or did I just pick a paper that did not have it in there? I am curious because maybe if the DC contributions were acknowledged in official releases to Scientists perhaps they could see the results possible and maybe more would jump on board and encourage others to do the same. Obviously 'getting the word out' is an important step to getting more and new users over time.

I think one way to do that would be to get more and more Scientists involved and aware causing them to use it in their vocabulary causing their students/whatevers to then do the same. This would then, hopefully, cause other students/whatevers to do the same bring DC into the home from the only people possible, the kids! As kids we all thought we could solve all the problems of the World if we just thought about it enough. As adults we begin to realize the daunting tasks involved. IF we could get the idea of DC being a positive influence and a venue for change maybe we could get it installed on every computer, or at least the Developed World ones! BUT I think if its involvement were acknowledged in official papers, it would start the thought processes flowing.

Very few papers, if any, mention DC. Every BOINC project has this "problem" from my point of view! It seems that Distributed Computing is still a "toy" to play with, something that "could" become important but now is not.
Really, the only project that I know which lives on DC is E@H (in fact you can see that the organization itself behind it is on top of boinc resources donors). Here on this forum, maybe one year ago, I read that rosetta@home lags behind many new features because BOINC is very different from a HPC, first and foremost because each core has to run independently, while many optimizations would require interactions between boinc nodes.

For Dr. Baker, why not update BOINC website page dedicated to papers? It could show your group's great papers to a much wider audience!