Talk:Folding@home

From Wikipedia, the free encyclopedia

Former FA Folding@home is a former featured article candidate. Please view the links under Article Milestones below to see why the nomination failed. For older candidates, please check the archive.

Contents

[edit] clarification needed

Recent developments in the protein folding world by both David Baker's research group at the University of Washington and Rama Ranganathan's research group at the University of Texas Southwestern Medical Center have demonstrated that the folding of proteins, as well as protein-protein interactions and protein engineering can be best solved using principles of cooperative evolutionary conservation. This is not the approach taken by Folding@home, which has rendered it somewhat obsolete.

On the "This is not the approach taken by Folding@home, which has rendered it somewhat obsolete." Does this mean Folding has rendered those paper's conclusions somewhat obsolete, or those papers have rendered the Folding project somewhat obsolete? —The preceding unsigned comment was added by 71.192.222.26 (talk • contribs) .

I understood it as the latter, but I am not really sure. There really should be a source to back up this claim, whatever the actual meaning is. --PS2pcGAMER (talk) 06:02, 14 February 2006 (UTC)

It is the latter. Both Rosetta@ and Folding@ find the minimum energy of protein structure (i.e. the correct fold). It does so by probing the energy landscape (conformation) and looking for those "moves" that decrease overall energy, therefore mimicking how nature folds the protein. So yes, they do more or less the same thing and Rosetta@ has obtained results that are far more impressive than Folding@. But Mr. Pande would have you think otherwise. I find it disturbing that the original text (above) was completely supressed from the article in favor for the "authority" in the field, the owner of Folding@home, Mr. Pande. I thought Wikipedia was more democratic and less authoritative, but I guess I was wrong. Cornosalpo 22:34, 6 April 2006 (UTC)

Cornosalpo, Mr. Pande post is perhaps not perfect (as we'd need and would like a response from Mr. Baker), but he is much more of an authoritative voice on this project than some random unknown person that changes the text without real substantiation. Comments, like yours, belong in the discussion, not on the article.

One way to look at it is: Rosetta works related to prediction of a protein's correct structure (detection and hopefully how to fix or nullify/destroy that protien). Some of the work is to improve its accuracy. F@H works to find out why/how a protein is misfolding (and hopefully how to prevent it to keep happening). These are two different approaches to medical research that do not directly compete against one another (as they seek different answers) - except for volunteers (CPU cycles). Both have made some progress but much work is still required.

Sources & references & qualifications please! Although I'm happy to see that the wiped part of the article by Kiio was reposted in this discussion area, it unfortunately made a contentious statement that had no supporting material or comment about qualifications to make such a statement regardless if it was correct or not. It shouldn't have been wiped out in the first place, rather just the contentious statement should have been moved to this discussion area. --ArkW5 14:40, 9 June 2006 (UTC)

Clarification by Mr. Vijay Pande http://forum.folding-community.org/viewtopic.php?p=125338#125338

I know Baker and Ranganathan and their work very well and (like the rest of the protein community) find their work very important and impressive. However, Rosetta and Folding@Home are addressing very different problems. Rosetta only predicts the final folded state, not how do proteins fold (and Rosetta has nothing to do with protein misfolding). Thus, those methods are not useful for the questions we're interested in and the diseases we're tackling (Alzheimer's Disease and other aggregation related diseases). Also, one should note that accurate computational protein structure prediction is still very challenging compared to what one can do experimentally, whereas the information obtained from Folding@home on the nature of folding and misfolding pathways matches experiment (eg with quantitative validation in rates, free energy, etc) and then goes beyond what experiment can tell us in that arena. While Rosetta has gone a long way and is a very impressive project, given the choice between a Rosetta predicted structure and a crystal structure, one would always chose the crystal structure. I bet that will be changing due to their great efforts, but that may still be a ways off for that dream to be realized. So, both are valuable projects IMHO, but addressing very different questions. I think there are some misunderstandings out there, though. Some people think FAH is all about structure prediction (which it is not -- that's Rosetta's strength) and some think Rosetta is about misfolding related disease (which it's not, that's Folding@Home's strength). Hopefully this post helps straighten some of that out.

To Ucarambo, I noticed you edited the above quote on the 26 April 2006 - I reverted it as it was a quote from another source. But so not to delete your comments (you probably didn't realize it was a quote), I've listed the changes you had made to the above quote below: "Rosetta and Folding@Home are not addressing very different problems." "Rosetta only predicts the final folded state, and how do proteins fold (and Rosetta has nothing to do with protein misfolding). Thus, those methods are useful for the questions we're interested in and the diseases we're tackling (Alzheimer's Disease and other aggregation related diseases)." "So, both are valuable projects IMHO, but addressing very similar questions." "and some think Rosetta is about misfolding related disease (which it is)" Thanks for your input. However, if you can provide some sources (e.g. a post from someone working with Baker Laboratory) that will be a big help. --ArkW5 14:40, 9 June 2006 (UTC)

In addition the above thread (that includes the above post from Dr. Pande), I also suggest looking at the the thread The Link between papers published and applications --ArkW5 14:40, 9 June 2006 (UTC)

[edit] Folding@Home and Cancer

First results from Folding@Home cancer project published. Posted by Dr. Pande (Jan, 2005): "We have been studying the p53 tumor surpressor and our first results have recently been published. To our knowledge, this is the first peer-reviewed results from a distributed computing project related to cancer. Thanks to the continued support of FAH donors, this is will be just the first of many cancer related works that will come from FAH. The nature of our results can best be described in our paper. However, here's a brief summary of our results. Roughly half of all known cancers result from mutations in p53. Our first work in the cancer area examines the tetramerization domain of p53. We predict how p53 folds and in doing so, we can predict which amino acid mutations would be relevant. When compared with experiments, our predictions have appeared to agree with experiment and give a new interpretation to existing data." "I should stress that cancer is a tough problem and no single paper will cure it (especially since there are so many different forms of cancer, and the term "cancer" is an umbrella term for them). However, I am very excited about the future direction of this work, especially as we start to couple it with the new drug design methods we have been developing." Paper is the 20th listed on Papers --ArkW5 14:40, 9 June 2006 (UTC)

[edit] More clarification

"So far, the Folding@home project has successfully simulated folding in the 5-10 microsecond range"

  • Someone please define this to the uninitiated.

I was hoping to find a section on results. I know there is a number of links on the Folding@home project page, but like the above comment, I don't really understand them. Could someone write or link to something written for an inteligent audience that is not already familar with the subject? TIA, Mike

I believe what they mean is, when a protein is folding, it takes a certain amount of time to do so in real life. This isn't referring to how long it takes to simulate on computers. Rather, it is an impressive figure because of the extent a protein can be accurately simulated today. I would think it's more a testament to the good programming than the processing power, however, you can't have one with out the other. CryptoQuick 09:38, 30 October 2005 (UTC)

[edit] references

I added references to the comment on "project significance".

However, it would be good to put an in-text ref number. I don't know how to do this. Can someone attach both ref papers to that paragraph?

—The preceding unsigned comment was added by Kiio (talk • contribs).

[edit] F@H on Wikimedia servers

I removed the following from the article:

On December 6, 2005, F@H was installed on some of Wikimedia's servers.[1]

It just seems very unencyclopedic (kind of fits in with Wikipedia:Avoid self-references) and also there are most likely quite a few companies that have F@H installed on various workstations or servers so it just seems unnotable. Furthermore, the section it was in doesn't really fit in either. I wanted it listed on the talk page though because it is interesting. PS2pcGAMER 00:38, 13 December 2005 (UTC)

[edit] F@H-on-BOINC

The article says that Folding@BOINC is currently under development. What is the official status of that venture? It seems to the very casual observer that progress has stagnated at the very least. Billy the Impaler 13:18, 18 July 2006 (UTC)

[edit] Commercial / non-commercial project?

The article doesn't really state whether this project is commercial or non-commercial. In other words, if some earth-shaking proteing thingie that can cure AIDS is found, will some commercial instance just shut down this project, say thanks and then make billions of money? Or will the cure for cancer be posted to public and then medical companies can compete, who can create the cheapest cure? --ZeroOne 01:24, 24 August 2006 (UTC)

F@H and the Pande Group has published all their results and makes them available to anyone vis the "papers" link on their homepage. They are a non-profit research organization. Link: http://folding.stanford.edu/papers.html Billy the Impaler 18:59, 1 September 2006 (UTC)

[edit] citation needed

"Folding@home is now the second largest distributed computing project after SETI@home."

It is not possible to compare different DC projects by size. Every project will use different methods to describe its size. Is it (active) CPU count? Is it GFLOPS count? etc. It is a mess.

The FAH DC system "size":
http://fah-web.stanford.edu/cgi-bin/main.py?qtype=osstats

BOINC DC system "size":
http://www.boincstats.com/
http://www.boincsynergy.com/stats/index.php

[edit] no connection to the FAH

Moved this from the Folding@home page as this has nothing to do with the FAH.

  • P. Bradley et al (2005). "Toward high-Resolution de Novo Structure Prediction for Small Proteins". Science 309 (16 Sep.): 1868–1871. 
  • M. Socolich et al (2005). "Evolutionary Information for Specifying a Protein Fold". Nature 437 (22 Sep.): 512–518. 

—The preceding unsigned comment was added by 194.204.35.117 (talk • contribs) 11 October 2006.

I confirmed those with Prof. Vijay Pande and those are indeed not about Folding@home.

[edit] Changing quotes

I do not think it is a proper thing to do to change quotes.

On 21:49, 2 November 2006 Bucetass changed Prof. Vijay Pande quote:
"Some people think FAH is all about structure prediction (which it is not -- that's Rosetta's strength) and some think Rosetta is about misfolding related disease (which it's not, that's Folding@Home's strength). Hopefully this post helps straighten some of that out."
to this:
"Some people think FAH is all about structure prediction (which it is not -- that's Rosetta's strength) and some think Rosetta is about misfolding related disease (which it is). Hopefully this post helps straighten some of that out."


Why was this needed? It just negates the meaning of the Prof. Vijay Pande statement and he is only person who can do it, not someone else.

[edit] "Progress" section

The "Progress" section relates only progress in building the computing community, the virtual supercomputer, not progress in achieving the scientific ends to which the technology is being applied. It would be useful for someone to add information on progress toward scientific ends. --User:Ceyockey (talk to me) 05:02, 26 November 2006 (UTC)

[edit] Graphical processing units

"However, it should be noted that this exaggerates the performance increase of the GPU client over the CPU client: the CPUs that contribute to Folding@Home vary widely from new to old, high performance to low, whereas the GPU client runs on only the very latest GPUs from ATI Technologies. A comparison to the latest Opteron processor showed more modest gains, both in terms of performance in points and points-per-watt.[1]"

This statement is just wrong because a TFLOPS count and WU points are not comparable. WU points are tied to WU completion time, not to system performance running the WU.


—The preceding unsigned comment was added by 194.204.35.117 (talk • contribs) 27 November 2006.

  • To my reading, this section seems to include reference to the same recent public beta test twice. Could someone familar with the beta test clarify/consolidate this please?

Stanford has recently cited further advances with the high performance client and stated they will be releasing a public, beta trial at the end of September 2006. ... As of October 2, 2006, the FAH GPU client has been released into a public beta test. Keesiewonder 01:26, 13 December 2006 (UTC)

[edit] Link to human gene

I added the dab notice at the beginning. I don't appreciate the tone of your edit note, but that's just an indication of your style. The red-link was valid, but I won't contest its removal. 'FAH' is the official symbol for the gene mentioned; I've added a request for article for the gene, as it is involved in human disease. Biologists who search 'FAH' will be frustrated by not finding the x-reference. Be that as it may, addition was a violation of the 'do not predict the future with edits' dictum. --User:Ceyockey (talk to me) 02:33, 4 December 2006 (UTC)

[edit] Deletion of 'see also' section

User:Records deleted the 'see also' section with the edit summary "Removed See Also as links in article direct to See Also articles, besides FA articles don't have see also (todays one)." It is patently incorrect that FA articles do not have See Also sections. Clicking through random articles at Wikipedia:Featured articles will illustrate this. --User:Ceyockey (talk to me) 02:37, 4 December 2006 (UTC)

[edit] Needed

This article could use some info on how they get people to install it without their knowledge. It happened to me and I did a little google0ing, and apparently getting this program on your computer without your knowledge is not unheard of. 66.231.130.14 23:41, 9 December 2006 (UTC)

This shouldn't be put into the article because it is against the folding@home license agreement. When you install FAH you are supposed to have permission to install it on the computer so installing it without permission/knowledge is not something appropriate for this article. SirGrant 01:10, 10 December 2006 (UTC)

A mention that there are instances of illicit installation without people's knowledge happening on the internet ... if verifiable and documented ... is not out of the question in the article. It's tantamount to adding an 'FAH as malware' section. However, it must have good documentation; otherwise, it comes across as just a snarky scare note intended to make the public wary of this but not other distributed computing efforts. --User:Ceyockey (talk to me) 01:16, 10 December 2006 (UTC)


F@H EULA Violation Removal Tools:
http://ra.vendomar.ee/~ivo/ufold/

—The preceding unsigned comment was added by 194.204.35.117 (talk • contribs) 28 December 2006.

[edit] More information which can be added to article

How does the Folding@home technique work:

Find at the project website, or in the project forums. Judging by the titles of published material which can be downloaded from the site, you may find answers amongst them: "Mathematical Foundations of ensemble dynamics.", "Atomistic protein folding simulations on the submillisecond timescale using worldwide distributed computing.", and "How well can simulation predict protein folding kinetics and thermodynamics?" are but some. --Foundby 11:44, 16 December 2006 (UTC)

[edit] Can we use one name throughout the article.

Folding@home, Folding@Home, and FAH are all used in this article. Can we just use one? To make it even more complicated, Stanford's own websites uses both capitalizations... BebopBob 02:17, 21 December 2006 (UTC)

Agreed, I think we shouled use Folding@home for the first time it is mentioned in each section and than F@H for each subsiquent reference. That way it would be constitant. The reason I think we should use Folding@home for the capitilization is because that is how it appears in tht title image on the website homepage. SirGrant 22:05, 22 January 2007 (UTC)

I have gone through this article and changed the first instance of F@H in every paragraph to Folding@home and all subsequent instances in that paragraph to F@H. Hope this helps with consistency. BebopBob 00:06, 5 March 2007 (UTC)

[edit] correct direction?

"the Folding@home project has successfully simulated folding in the 5-10 microsecond range—a time scale thousands of times longer than was previously thought possible" I am guessing they meant to say thousands of times SHORTER here, as they are outlining its successes. If you know that's what they meant to say, please change. Jwigton 05:27, 13 January 2007 (UTC)

Longer is correct. These simulations took months/years/decades of processor time in order to simulate a few microseconds of the real-life folding of a small protein. Previously, if we take the contents of that quotation to be accurate, only nanosecond simulations had been done. Thus, we can see 1000x more of the folding pathway than we could have before. --Antelan 05:17, 15 January 2007 (UTC)

[edit] Folding@Home teams sounds like an ad?

I was reading through this - the section "Folding@Home teams" sounds like an advertisement, or at the very least, not like an encyclopedia... should probably be fixed, or tagged with some template (i'm not sure which would be appropriate here). Fiskars007 18:16, 14 January 2007 (UTC)

I removed some stuff that can't be verified, but the paragraph still puts a somewhat positive spin on the benefits of being on a team.--Planetary 00:34, 15 January 2007 (UTC)
I added a NPOV template.--Donald Goldberg 01:59, 11 February 2007 (UTC)

Fine, but what is the NPOV dispute here?

—The preceding unsigned comment was added by 194.204.35.117 (talk • contribs) 13 February 2007.

I don't think there's an NPOV dispute here, certainly not with the whole article, so I removed the NPOV template and added the advertisement template to the "Folding@home teams" section. 151.204.22.18 20:31, 25 February 2007 (UTC) (User:Daekharel, not logged in)

Agreed I don't think it is a problem with the article in general just that section that needs to be reworked SirGrant 21:46, 25 February 2007 (UTC)
I rewrote it. I think it sounds better now. BebopBob 23:43, 4 March 2007 (UTC)

[edit] FAH NPOV dispute

Is there a FAH NPOV dispute? 194.204.35.117 10:19, 20 February 2007 (UTC)

[edit] Article improvements needed.

I'm probably going to offend some people who obviously worked very hard on this article. So let me apologize in advance. Please take these comments in the spirit in which they are offered which is for the betterment of wikipedia and F@H. In no particular order...

- If I didn't already know most of this information this article would be very difficult to parse and understand.

- There is virtually no flow between the sections. One section should lead naturally into another. I know this is difficult in technical writing but it can be and should be done.

-There is insufficient description of terms and acronyms for somebody who is reading about F@H for the first time.

-The sections are out of order in terms of importance. The most important information should be first and so on.

-The advertisement tag is absolutely correct. This is probably always going to be somewhat of an issue because anybody who has the knowledge and interest to write about F@H is going to be somewhat biased. However, I think we can put F@H's best foot forward without overtly advertising.

-Citations and sources needed. There have been a ton of articles written about F@H. This is supposed to be the modern encyclopedia entry for F@H so cite those articles. I know the temptation for someone who has been involved in folding for awhile is just to say, "I know all of this stuff so why should I find a source or citation?" However for the article to be professional it needs to cite sources. This is the Wikipedia standard and it is there for a reason.

-Better description of the science that has been published so far. This works well as a subsection. Should include a chronological list of all results that Pande group has published with a short abstract of the article.

-Description of the team and points elements of F@H in a more thorough and unbiased fashion.

-Better linkage to F@H sites and resources with better description of the links.

-The two diagrams about relative computing power of various operating systems are great. However there should be some descriptive text attached to them to make them a little easier to grok for newcomers who are hitting these concepts for the first time.

I'm sure I'll think of more on 2nd, 3rd, and .... read throughs. However, I think this is enough to chew on for now. I'm willing to tackle some of this but I think there are more gifted and knowledgeable writers in the F@H community who could do much better than I. Rogsmart 08:37, 26 February 2007 (UTC)


[edit] Computational power image, please update

Now that PlayStation 3 is involved (and accounting for almost 5 times of Windows TFLOPs, and almost three-fourths of the operation) with this project, it needs to be added for the TFLOP image.
Windows: 152
Mac OS X/Power PC: 7
Mac OS X/Intel: 8
Linux: 43
GPU: 41
Playstation 3: 435

If you are updating the image, and the above figures are obsolete, please use current figures.

Socby19 17:52, 23 March 2007 (UTC)SocBy19


Given the huge differences between "late" March 23rd and "early" March 24th (PS3 total is now up to 577) I think it might be a good idea to wait until the numbers level off before further changes to the main page. Maybe it should be tagged as news, to note that the numbers are changing rapidly? Shadowstar 16:37, 24 March 2007 (UTC)

From the source, P3's contributions are now at 732 TFLOPs. The total TFLOPs for the project are now 988. That's getting damn close to 1 PFLOP. Antelan talk 18:42, 25 March 2007 (UTC)
And from the same source, the PS3s contributions have now dropped to 576 TFLOPs, with CPUs timing out. It was at 990 for a while. The sustained level should probably be left lower than the peak level later in the article. --Shadowstar 20:15, 25 March 2007 (UTC)
Agreed. Good call, whoever put the 'current event' tag on top of this page. It will probably take some time to see "normal" patterns emerging (since at this moment we can't tell what normal levels will be). Antelan talk 23:19, 25 March 2007 (UTC)

[edit] Playstation 3 section

67.160.51.232, please add a reference to your disputation of the numbers provided by Folding@Home's website. Shadowstar 01:07, 25 March 2007 (UTC)

This is POV of the author and has no place in the article without a proper reference and argumentation.

"Criticism of the accuracy of this number has been raised. Cell processor floating point operations, which are 7 of the 8 processors in a PS3, are single precision operations[2] instead of double precision FLOPs that PCs use. Double precision operations are more accurate and take longer to calculate. Critics argue that the "comparable" computing power being added to the Folding@home project by PS3's is about 1/10th of what is reported on the Stanford website(citation needed). Stanford's client statistics page claims to take a conservative approach to FLOP calculation, which may also contribute to this criticism. It is unknown whether or not Stanford is taking the precision of operations into account when calculating the FLOPs from the software cores."

194.204.35.117 08:14, 25 March 2007 (UTC)

Actually, that was my attempt at changing it to NPOV (I guess it didn't work). Someone else might want to try fixing it. Here's the original:

"As wonderful as these numbers are they should not be used to directly compare Folding@home with any supercomputer, nor should they be used to compare the contribution of PS3's to the Folding@home project with the contribution of regular PC's. The reason for this is simple, the PS3 Floating Point Operations are single precision operations instead of the double precision FLOPS that all PC's use. Double precision operations are much more accurate and take much longer to calculate, and so the PS3 is only estimated to perform double precision FLOPS at about 1/10th it's measured TFLOP speed. Therefore the actual "comparable" computing power being added to the Folding@home project by PS3's is about 1/10th of that indicated by the PS3's TFLOPS measurement. Factual reality is also that even if 42,000 PS3's were connected together to form a single supercomputer they might not even be allowed on the top 500 supercomputer list because of this issue, and if they were it would be as having 100 TFLOPS, not 1 PetaFLOP."

(Seriously, I can't imagine why people would imply Stanford is lying for 72 TFLOPS.) Shadowstar 15:03, 25 March 2007 (UTC)

Majority of FAH projects are running off the GROMACS core and thus the majority of FAH calculations are single precision. http://fahwiki.net/index.php/Cores 194.204.35.117 15:59, 25 March 2007 (UTC)

[edit] "Performance weighting"

"A Peak output of the project at 990 teraFLOPS was achieved on Sunday March 25, 2007 at which time a new weighting system for PS3 results was implemented that reduced the speed rating of those machines by 50%. This had the effect of bumping down the overall project speed to the mid 700 range and increasing the number of active PS3's required to achieve a petaFLOP level to around 60,000. Lately, the console accounts for nearly 2/3 of all teraflops."

This wasn't the case. There was no performance measurement adjustments made. As the active PS3/CPU/GPU/... counting is defined by "return of a WU within last X days" then a single WU within this time window will increase active PS3 count. Even if the active PS3 count is still going up then large portion of those finish only few/one WUs during the "active" period. As long as there are relatively small number of dedicated PS3 running the FAH client then we'll continue to see the TFLOPS rating fluctuations.

http://forum.folding-community.org/fpost174824.html#174824

194.204.35.117 13:21, 26 March 2007 (UTC)

[edit] Criticism?

Following section in it present form has no place on the FAH pages:

"Folding@home has recieved some criticism, notably for its power consumption. [2]"

If this is a criticism about anything then it is not about the FAH.

194.204.35.117 19:51, 6 April 2007 (UTC)

Retrieved from "http://en.wikipedia.org../../../f/o/l/Talk%7EFolding%40home_fed2.html"