Hi all. I am some doubts over picking model. Thank you for your time.
Since I have to model two 12 aa loop regions, many loopmodels have to be generated (500 or 1000 as recommanded by others). I did a set of 100 and a set of 500. I used the DOPE score to compare them and the difference between the models with the minimum and maximum scores were around 1000 in both cases. So I just wonder, are you all using the lowest DOPE score to pick the best model? If so, will generated a set of 100 loopmodels be enough instead of 500 or a 1000?
I shall also use the model for MD, so in this case, do I need to build so many loopmodels? Will MD suffice to explore the conformations? I am sorry for not understanding the concept of generating many loopmodels. Is the for "correctness" or exploring all the conformations?
Thank you for reading this and sorry if my questions sound too inexperience. I am still learning and thank you very much for any help.
(Also, I looked at profiles but sometimes the second best model seems to have better profile at the loop regions than the one with the lowest score, any other ways tthat you use to pick model?)
--------------------------------- Do you Yahoo!? Everyone is raving about the all-new Yahoo! Mail.
Hello, I generated 500 loops for two segments - each one 8 residues long.I choose the best 20 loops - the lowest DOPE. A DOPE difference between thehighest energy loop and the lowest energy one is 130.My questions are: 1) Can I simply choose the lowest energy loops as a native loops ? 2) Can I choose my best loops in a following way: - I assign the template loops as a reference structures - I calculate local RMSD (N, Calfa, C, O - only on the loop) for the besttwenty loops- I plot these data with the DOPE variations and I check a correlationbetween energy and structural variability 3) Can I choose my best loops in a following way: - I assign the lowest DOPE loops as a reference structures rest of a procedure like in point 2 4) What is an acceptable (aproximately) DOPE difference in a set of the best20 (30) loops between the higest energy loop and the lowest one ? 5) How many stem residues modeller algorithm moves during optimization ? 5) Maybe there is a different way to increase loop credibility in blindpredictions ? I am soory for my long list but I am really interested in this problem. Thank you ! Karol Kaszuba