Folding Forum

By that I mean, is it just in the stages of continually perfecting algorithms to better model protein folding? Since right now the best structures would be in Swissprot and experimentally derived, and those are considered better.

Or, are you starting to look at novel proteins relevent to research going on? Like say I found a rare transcript in my cDNA library of Alzheimer's brain tissue, and I found a putative full transcript in an assembled EST database like Unigene, and I used phage display and made some to confirm the sequence, would I maybe in the future be able to submit a project to folding @ home if there was promise of it being worth something, and have you been doing anything like that yet?

You may not understand the purpose of FAH. The primary purpose is NOT to predict structures. The primary purpose is to understand the folding process including all of the steps along the way.

Experimentally determining that there are two stable structures of a protein (e.g.- one of which is directly related to Alzheimer's Disease) does nothing to explain why a healthy person produces only the "good" form or to explain WHY/HOW the "bad" form is created.

Part of FAH, of course, is is to perfecting algorithms to better model protein folding, but lots is being learned from applying those algorithms to real cases of mis-folded proteins. Have you read the scientific papers? http://folding.stanford.edu/English/Papers

jheil wrote:By that I mean, is it just in the stages of continually perfecting algorithms to better model protein folding? Since right now the best structures would be in Swissprot and experimentally derived, and those are considered better.

Or, are you starting to look at novel proteins relevent to research going on? Like say I found a rare transcript in my cDNA library of Alzheimer's brain tissue, and I found a putative full transcript in an assembled EST database like Unigene, and I used phage display and made some to confirm the sequence, would I maybe in the future be able to submit a project to folding @ home if there was promise of it being worth something, and have you been doing anything like that yet?

If you wanted to find the structure from your gene, your best bet (in silico) would be homology modelling, and Robetta. Of course if you can express the protein and crystallise it, X-ray xtallography would be better (but it sounds like you already know this).

Yes, x-ray crystallography is very complex and time consuming, and you have to be an expert in that to actually do it.

Frankly I haven't one paper all the way through. I have read bits and pieces, and I will probably get to the wiki sooner or later. Some of the stuff in the peer reviewd articles is a little bit over my head as of yet. I'm just an undergrad and I take mostly molecular and biotech courses. Some of the biochem stuff in requires a certain degree of brain wracking. So, I'm just trying to get a better understanding here, which is sometimes more fun. I don't need the detail right away and posting references that might help with my question is always welcome. I'm not trying to waste any body's time. I just want to understand it to the best of my ability right now and I don't have the time to do all of the work of reading the articles.

I did take Bioinformatics right now and I realize homology modeling would be great for my example. A monte carlo method might work too, just to compare to the homology model.

Anyway. So FAH is not just a super computer put in place to crunch protein sequences and spit out structures for people who want them. It is like a project to better understand the variables involved in the process. That's tough, especially trying to understand it in different hosts. For example, a bacterial host will have a different pH in the environment where the ribosomes are churning out the proteins and different potential for oxidation depending on the hosts cytoplasmic environment. Also, does the protein contain a signal peptide and does it get transmitted into the periplasm where disulphide bonds might be helped to form. And that's just in "simple" procharyotic organisms, a mamalian system is much less well understood. Quite an ambitious project. I might have to get into those articles sooner or later.

A protein might fold differently due to a mutation in where properties of other proteins change, altering the protein of interest's folding environment. There are a million other things too... RNA transcript editing, changing sequence... viruses inducing RNAi to the detriment coincidently knocking out some cell function regulating folding. Those are just some far out things I can think of off the top of my head. Prions, I can't for the life of me conceive of how some stray proteins can be infectious.


There are current theories that have to do with protein folding, like ramchandran plots and hydrophobicity and certain orders of residues that will likely form an alpha helix (related to the ramachandran plot), so folding at home is just going down the road of refining the understanding of the process itself eh? Interesting.

Anyway... I should really go into research someday if this is making me this excited. I am thinking about grad school, we'll see.