Protein Folding Conference (F@h and experiments)

[Jesse_V]

http://folding.typepad.com/news/2012/06 ... rence.html
Personally, I this is one of the top blog posts written so far:

Guest post: Dr. Greg Bowman, UC Berkeley

We just had a protein folding conference at Stony Brook University in New York that was extremely encouraging. Both the experimental and theoretical communities are very excited about the results we are generating with Folding@home. In particular, they are excited about (i) our increasing ability to make quantitative connections with experiments and (ii) the long timescale dynamics for large proteins we are now able to capture. For example, we recently succeeded in folding an 80-residue protein on 10 millisecond timescales (paper is here). For reference, that’s about twice as many residues and about 1,000 times longer timescales than what most anybody else is able to achieve! There are now multiple experimental groups who are asking us to make predictions for them to test. So, we appreciate all your help and have plenty of new calculations for you to contribute to.

[Jonazz]

A common criticism I hear is that Folding@Home isn't taken seriously by the scientific community. This proves them wrong!

[Jesse_V]

A common criticism I hear is that Folding@Home isn't taken seriously by the scientific community. This proves them wrong!

I've heard that a few times. Another one was that the electrical usage was not worth the effort. AFAIK, neither have had much credibility behind them. Posts like these and other similar writings really negate those criticisms. It seems to me that such information ought be more prominent.

[Meh_Lay_Lay]

Now I may understand that some people do not consider the power usage worth it. But how would F@H not be taken seriously? I thought all these calculations were essential to the understanding of proteins or whatsoever.

[csvanefalk]

Now I may understand that some people do not consider the power usage worth it. But how would F@H not be taken seriously? I thought all these calculations were essential to the understanding of proteins or whatsoever.

As I understand it (and someone correct me if I am wrong), much skepticism is grounded in the claim that simulations do not hold the same value as physical observation. It seems that this is not a very strong case as far as the validity of the progress is concerned - simulation and observation go hand in hand in order to understand the complexity of protein folding.

[Jesse_V]

Now I may understand that some people do not consider the power usage worth it. But how would F@H not be taken seriously? I thought all these calculations were essential to the understanding of proteins or whatsoever.

As I understand it (and someone correct me if I am wrong), much skepticism is grounded in the claim that simulations do not hold the same value as physical observation. It seems that this is not a very strong case as far as the validity of the progress is concerned - simulation and observation go hand in hand in order to understand the complexity of protein folding.

Haters gunna hate. There are other sources of computational power that's used for protein folding, but AFAIK, Folding@home is the most powerful. But yes, there are some very strong statements verifying the fact that simulations and experiments are needed together.

Taming the Complexity of Protein Folding:

Solving the folding problem will ultimately require a combination of theory and experiment, with theoretical models providing a comprehensive view of folding and experiments grounding these models in reality. Here we review progress towards this goal over the past decade, with an emphasis on recent theoretical advances that are empowering chemically detailed models of folding and the new results these technologies are providing. In particular, we discuss new insights made possible by Markov state models (MSMs), including the role of non-native contacts and the hub-like character of protein folded states.

Atomistic Folding Simulations of the Five-Helix Bundle Protein λ6−85

Solving the folding problem will ultimately require a combination of theory, simulation, and experiment, with theory and simulation providing an atomically detailed picture of both the thermodynamics and kinetics of folding and experimental tests grounding these models in reality. However, theory and simulation generally fall orders of magnitude short of biologically relevant time scales. Here we report significant progress toward closing this gap: an atomistic model of the folding of an 80-residue fragment of the λ repressor protein with explicit solvent that captures dynamics on a 10 milliseconds time scale.

[bruce]

As I understand it (and someone correct me if I am wrong), much skepticism is grounded in the claim that simulations do not hold the same value as physical observation. It seems that this is not a very strong case as far as the validity of the progress is concerned - simulation and observation go hand in hand in order to understand the complexity of protein folding.

I'm not a professional biochemist but in my profession (aerospace) but have value. When manufacturing produces a number of "identical" parts, they're actually not quite identical. How many does my company have to overstress to prove every airplane will be able to fly safely for the lifetime of the aircraft? Certainly some parts DO have to be intentionally broken in every conceivable way, but there's also a lot of calculations required to demonstrate that some conditions that can't be tested are also safe. Then, too, simulations can teach the engineers WHY and HOW things worked correctly and HOW or WHY they didn't work correctly. What good is the in vivo test of an Alzheimer's protein where you observe the patient died if you don't know WHY or HOW the clumps formed and if you can't observe the minute details that could very well be important -- and which are clearly visible In silico.

Best example I can think of: Something in the simulation causes patient death. OK, now run the simulation clock backwards and simulate it un-happening until you can figure out the critical step that went wrong. (This works better in deterministic models than in stochastic models, of course.)

[VijayPande]

A common criticism I hear is that Folding@Home isn't taken seriously by the scientific community. This proves them wrong!

Here's some facts that might be useful for talking to people who have doubts. By every metric used in the scientific community, we do really well:

* citations: I have an h-index of 55 with over 10,000 citations, primarily due to our work with FAH.

* awards: due to FAH, my group and I have been awarded key prizes. Most notably, I'm only one of two people to win both the Protein Society Young Investigator Award and The Biophysical Society Young Investigator award --- two key prizes from the scientific community based on our work with Folding@home. There are numerous other awards.

* grants: we have been very successful in peer reviewed grants.

* papers: our papers are in key, peer reviewed scientific journals

* conferences: my team and I present our work at invited talks at key conferences (I'm at one right now in fact)

With that said, there will be people skeptical about new methods, but I find that these are people *not* in our field (eg not in protein folding) and not familiar with our work. I could see how someone unfamiliar with our work would be skeptical (it probably sounds crazy that we could pull off what we've done, brining millions of people together to make scientific advances). If you encounter such people, probably sending them to the page which talks about our awards (http://folding.stanford.edu/English/Awards), papers (http://folding.stanford.edu/English/Papers), or citation impact (http://scholar.google.com/citations?use ... AAAJ&hl=en) is a good place to start.

Finally, it's worth pointing out that our key predictions are tested experimentally, either by my lab or others. Our work on Alzheimer's Disease recently published in J Med Chem is primarily an experimental test of previous FAH simulations some years ago.

[verlyol]

Effectively there are always people who are skeptical with new research methods, this is not new and it has always existed before.

In a sense Folding@home is a victim of its immense success and reputation of most powerful distributed computing network.

Just as the saying Prof.Pande, if we encounter skeptical people , we must invite them to visit the page of awards ... it's an easy way to stop the skeptical comments

[Jesse_V]

Dr. Pande, that sir is very impressive.

I didn't know what an h-index was, so I looked it up, and for those of you who don't know, it means that he has written 55 scientific papers which have been each cited by other papers at least 55 times. The Google Scholar page also gives a i10-index of 145, which means that 145 of his papers have been cited by others at least 10 times. And that truly is remarkable. David E. Shaw, the head of the group that runs the Anton supercomputer, (the 2nd most powerful system for studying protein folding) apparently has an h-index of 16, according to Microsoft's Academic Search.

It seems that Prof. Charalambos Kalodimos of Rutgers University has also won both the awards. I assume this is the other person then. Also impressive.

Anyway, I think that's about as strong a rebuttal to the criticism as it could get! "...Cave Johnson, we're done here."

[csvanefalk]

The moral of the story:

Never. Stop. Folding.

[moritzgedig]

In: http://folding.typepad.com/news/2012/06 ... -come.html
Prof. Pande looks back and extrapolates.
I do not see how the growth could be sustainable.
In the past there were growth mechanisms added (*):

• Regular hardware development
• Additional Users
• New Hardware types(GPU, PS) *
• Maturing Software
• New methods

What will add performance at the level of the past? Isn't saturation/stagnation more realistic?
The Project seems to have come to a point were the effort can be shifted from methods to results.

[7im]

Growth is easy when less than 1% of GPU owners are folding.

And the number of CPU and GPU cores per system keeps doubling.

The project has also released the new V7 software that is much easier to install and run, attracting a larger number of more general computer users.

And methods vs results are not so disimiliar that the project needs to shift between them. I see lot's of potential...

[VijayPande]

We have plans for growth that could easily see 2x to 10x growth in a few years. Just as the idea of folding on GPUs and PS3s probably seemed like a pipedream in 2003, our plans now are pretty ambitious and (I think) ground breaking. I'm playing our cards a bit close to our vest here (sorry), especially since I hate the idea of vapor ware (where we announce something but it takes forever for it to come out), so I'm not being specific intentionally (sorry), other than saying that this is very much on our mind and we are actively pursuing new directions.

[Jesse_V]

We have plans for growth that could easily see 2x to 10x growth in a few years. Just as the idea of folding on GPUs and PS3s probably seemed like a pipedream in 2003, our plans now are pretty ambitious and (I think) ground breaking. I'm playing our cards a bit close to our vest here (sorry), especially since I hate the idea of vapor ware (where we announce something but it takes forever for it to come out), so I'm not being specific intentionally (sorry), other than saying that this is very much on our mind and we are actively pursuing new directions.

F@h was all over the news when GPU and PS3 folding was made possible, and we saw spikes in participation. I don't mind you being coy, because it looks like you're actively moving in good directions. I'm looking forward to seeing what's next...

Computing: _____ of the world, unite!