I generally understand what the Rosetta@Home project is all about (understanding and predicting how proteins fold and interact with each other), but I have a question. Is there a point at which the project will be considered complete? Would it be when researchers can both synthesize proteins exactly the way they want them and predict how proteins will interact based on known "rules"?

I generally understand what the Rosetta@Home project is all about (understanding and predicting how proteins fold and interact with each other), but I have a question. Is there a point at which the project will be considered complete? Would it be when researchers can both synthesize proteins exactly the way they want them and predict how proteins will interact based on known "rules"?

Rosetta doesn't study how proteins fold per se, basically what it does it try and find the native structure by MSMs, Rosetta doesn't know why that structure is native. Therefore, since Rosetta is simply a tool for protein structure prediction, as long as people haven't solved the folding code, there will always be structures to predict. Defining rules for protein folding is so complex, Rosetta doesn't really look at this, a better example would be folding@home which stimulates long time scale folding of a protein of known structure. Further, until the day where you can input a linear sequence and press a button and the native structure is solved and you can simply print out the protein, Rosetta will still be useful for predictions.

Hi CM

To expand on what Carl has said, while Folding@home attempts to model the process of protein folding, Rosetta attempts to determine the final structure that the unfolded protein (so the string(s) of amino-acids and any cofactors) will take, which is computationally vastly quicker than trying to determine the shape by calculating each step of the folding process, although it is still an incredibly large calculation.

How far Rosetta (or any other protein modelling software) is from perfection I don't know, but the bigger the pool of computer power available to the researchers, the faster it will get there. In reality, my guess is that it will get to the point where it's "good enough" for some tasks and types of proteins before it is at that point for others, and will already be beyond that point for many cases.

In addition, Rosetta is used for protein design - that's modification of existing proteins and the design of completely new ones to perform required tasks. That's an incredible step and opens up so many possibilities.

So to answer your question, is there a point where it will be perfect?: possibly, but that's still far enough in the future to be sci-fi. The reality in the foreseeable future is that it will be good enough for a larger and larger pool of tasks, and will keep getting better. The significance of that pool of tasks shouldn't be understated though as proteins are fundamental to a huge range of disciplines, many of which can have a massive positive effect on our lives, from helping to understand and cure illnesses to reducing pollution.

Danny

---

I don't have the basic knowledge to understand everything you both wrote, but your replies prompted me to look up "MSM" - which led me to "stochastic process" - and "native state" so I have a little better understanding of how ignorant I am. Thanks. LOL. Seriously, though, thanks for taking the time to reply.

Hi CM,

In the Baker laboratory at UW in Seattle (aka institute of protein design) we focus on the design of new proteins with new functions, modification of existing proteins to enhance activity and for small proteins that can bind to battle disease.

Rosetta@Home is a calculation (not a folding simulation) that can help us to discriminate between good and bad designs. It uses naturally occurring proteins as an empirical/stochastic source. Thus, the calculations tells us what designs are more likely to be produced by bacteria/yeast than others. So it's kind of an extended probability calculation.

As long as there are targets and problems to be solved for which proteins can be reasonably used, Rosetta@Home will be used by researchers around the globe.

Best,
Ruud van Deursen