1)I recommend that, in order to attract and retain more CPU time contributors, more thought be given to over all web site presentation. Not fancy bells and whistles just improved more efficient presentation WITH just a touch better sales effort.

SPECIFICALLY
2)I recommend that the ABOUT page - the first explanation paragraph - be broken into at least three separate paragraphs for better understanding.

3) I recommend that the Work Unit Graph showing all the protein bending, folding, and mutilating include:

A) A Legend as to what the colors of the different sections of the protein are and
B) What low and high energy states mean and
C) Other helpful info, briefly, right there on the graphic.

In this way, when a contributor brags to a friend about Rosetta, he can go straight to the blinking lights and moving images and his friend can say, "Hey man! that's really cool. I understand it and I want to help"

It would be more rewarding and meaningful to Contributors to not require jumping around to several different areas.

Stacey Baird - Davao PhilAmericans, Davao City Philippines

---

Yes, explaining leading edge scientific discovery has proved challenging.

One explaination I posted just got a vote for a sticky because someone liked it that much. Thought you might want to have this in your bag of tricks when you are trying to explain it to people.

Adding to your idea a bit... Perhaps a more interactive graphic? A way to actually click an area of the graphic and have it bring up some help text about what it's purpose is, or what type of information is displayed in that area.

I for one would love the ability to actually SLOW it... just for a few minutes while I explain it to someone. And also a way to zoom in to the atomic level and see the twists and turns and chemical bonds... again, while in slow mode. In fact, a "slow mode" would be a possible way to run a cooler, <100% CPU if people wanted... naw I'll just wait for the BOINC changes that are coming... I want "slow mode" to be like, 1/20th of live.

---

Add this signature to your EMail:
Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might!
https://boinc.bakerlab.org/rosetta/

Yes, explaining leading edge scientific discovery has proved challenging.

One explaination I posted just got a vote for a sticky because someone liked it that much. Thought you might want to have this in your bag of tricks when you are trying to explain it to people.

Adding to your idea a bit... Perhaps a more interactive graphic? A way to actually click an area of the graphic and have it bring up some help text about what it's purpose is, or what type of information is displayed in that area.

I for one would love the ability to actually SLOW it... just for a few minutes while I explain it to someone. And also a way to zoom in to the atomic level and see the twists and turns and chemical bonds... again, while in slow mode. In fact, a "slow mode" would be a possible way to run a cooler, <100% CPU if people wanted... naw I'll just wait for the BOINC changes that are coming... I want "slow mode" to be like, 1/20th of live.

---

Thanks for the Soda Straw analogy. It does help. What about the low and high energy states - are these static charges or net charges such that the like charges repel and the net charge of the group of proteins + - + - = 0 and must equal 0 in order to bind together - or something?

---

Thanks for the Soda Straw analogy. It does help. What about the low and high energy states - are these static charges or net charges such that the like charges repel and the net charge of the group of proteins + - + - = 0 and must equal 0 in order to bind together - or something?

The side chain of a protein can have different areas that will make bonds with one another:

- "fat-like" parts that will aggregate and repell water. These are uncharged.

- different residues like amino groups (NH2), imino groups (NH), acid groups (COOH), that can accept or donate protons depending on the pH of the solution. These will yield positive (NH3+, NH2+) or negative (COO-) charges and then be able to attract groups of opposite charge

- Residues that can form hydrogen bonds. This can take place between OH, NH, NH2, COOH, groups and is a form of binding forces between polarized but uncharged groups which can also be bound together by adding water between them (so called water mediated hydrogen bonds) [edit: It is a little more complicated as charged residues may also form hydrogen bonds, but I think one should not make it too complicated, the model should just account for the three principal bonding mechanisms.]

One might realize such interactions by using magnets (best would be ring shaped magnets with just one pole around them, cause only then will the same part repell an opposite pole in all orientations) for charges and small areas of hook-and-loop-fasteners for hydrogen bonds.

I still have no clue how to realize the fat-like parts. Any suggestions? How do you realize a region without charge and without polarization that still does stick together simply by van-der-Waals forces and the fact that it repels water?

Maybe one should only have a single sophisticated model made of some kind of garden hose and a number of smaller models with just a three-color coded attraction: probably press studs like on rain jackets on them and you may only connect parts of the same color?

You would present the big model and explain the folding principle. Then you hand out the smaller ones (some kind of jointed plastic tubes that can rotate aorund the joint - is there a toy like that that allows for making long chains?)

It'll show the principle in protein folding in a true hands-on experience:

- the more you stick to connecting the adjacent parts first the more compact and lower in energy the protein shape will become in the end;
- you will also see that the number of conformations is vastly reduced if you stick to that "regional bonding comes first" principle.

In the end all participants may compare their shapes and see that there still are no two alike in spite of the reduction of possible conformations.

What about the low and high energy states??

In a word, lowest is best. A large negative number indicates strong chemical bonds throughout the protein chain. And this tends to be found in the native structure of the protein, so low energy readings tend to be the closest matches.

The other thing you asked about was the colors. (Yeah! Why doesn't the description of the graphics explain that???) I believe the color is just a frame of reference. Start on one end of the protein with red, and the other end with blue and work your way through to the middle, shifting the colors slightly as you go. I believe it is just a frame of reference to help the eye follow the strand through the twists.

And if you click and drag the mouse over the low energy structure show in the graphic, you can rotate it around and look at it from all sides. This is most useful when we're comparing Rosetta's prediction with a known structure, because you can rotate the known structure into the same orientation as your prediction and see how close they look (which CASP, we're modelling unknown structures, so there is no "native structure" known to compare to). You can also right click and drag up and down to magnify and reduce the low energy structure.

Welcome aboard! Crunch more Rosetta!

---

Add this signature to your EMail:
Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might!
https://boinc.bakerlab.org/rosetta/

Mr. Baird is right on track with his message...

As I've said before, the general appeal or 'outreach' should be geared toward people that never ever read these forums. They are people that wind up saying 'wow, ok, I can do that on my computers....thanks for showing me"

It's an example of 'Keep it simple, stupid'. :-)

---

Founder of BOINC GROUP - Objectivists - Philosophically minded rational data crunchers.

What do you think? Are you saying you feel the graphics should be changed to be more understandable to the lay-person?

For example, perhaps rather than showing an "energy level" they could show a "model score"? And have it be on a scale of 0 to 100 (or 1000?) or something less confusing then positive and negative numbers?

I for one would like to see a more consistent scale on the energy shown. For example, if I'm currently crunching model 20 and haven't seen any low energy structures yet, then I can't see my prior 20 predictions, because the scale on the graphic hasn't included them yet. It would seem as though the starting scale of the graph should be at least as low as the lowest model I've crunched. That way I could always see my prior predictions in the graph... and how far away from them I am.

I understand the energy range is different for different proteins and models and techniques, and in fact isn't actually known ahead of time (unless crunching a known protein)... but you've got to be able to establish an outside range based on the number of amino acids (AAs) perhaps? I mean I don't think I've seen a negative energy that exceeds the number of AAs. So, at the outset perhaps a scale of -AA thru +AA would encompass all the possibilities? (or perhaps +/- 2*AA?) And then scale that in to a 0 - 1000 scoring? Where 1000 is the best (i.e. most negative energy) model.

What do you think? Would that be helpful to us lay-people to "understanding" what's going on?? (vote "yes that would be helpful" with the green plus sign under this post).

---

Add this signature to your EMail:
Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might!
https://boinc.bakerlab.org/rosetta/