I must admit I've been dirtying up the "Comments/questions on Rosetta@home journal" thread so I'll stop, sorry :oops:.

Does the full atom relax algorithym, time slice ie. move the biggest attraction/repultion a very small amount then look for the next biggest attraction/repultion, or just go for the biggest attraction / repultion join/ repulse that bit then go on to the next biggest attraction/repultion etc... I must admit it's puzzeling me that a ab-inito structure is so close to the RMSD and yet is not being pulled into shape by the full atom relax. I would expect it really to be made the right shape, I would of though that close enough is literaly close enough for the full atom relax.

Edit too make a bit more sence

---

I must admit I've been dirtying up the "Comments/questions on Rosetta@home journal" thread so I'll stop, sorry :oops:.

Does the full atom relax algorithym, time slice ie. move the biggest attraction/repultion a very small amount then look for the next biggest attraction/repultion, or just go for the biggest attraction / repultion join/ repulse that bit then go on to the next biggest attraction/repultion etc... I must admit it's puzzeling me that a ab-inito structure is so close to the RMSD and yet is not being pulled into shape by the full atom relax. I would expect it really to be made the right shape, I would of though that close enough is literaly close enough for the full atom relax.

Edit too make a bit more sence

this is roughly what we do. but the landscape is extremely bumpy since small changes in atom positions can cause huge clashes. so it is possible to start pretty close to the deep native minimum and still not fall in. we're still working to improve this part of the search

---