[modeller_usage] What is a difference between automodel and loopmodel with respect to refining only part of the model optimization ?

Dear All,

Let me explain that what I have tried:

I have generated 100 models for one single point mutation sequence following optimization(loop model) on single point mutated residue along with neighbour residues(with different sphere range 1A°-10A°). Now, the observation from my out put is as following:

1. When i optimized the SPM model with the large sphere (10A°) of neighbour residues, I got bad model compared to crysal structure but DOPE score was really high(~5K).

2.When i optimized the SPM model with the less sphere (1A°) of neighbour residues, I got similar model compared to crysal structure but DOPE score was really bad(~250).

So , I dont understand why I am getting less DOPE score when I have similar good models compared to crystal structure and vice versa.

Then, I followed the same procedure but automodel default optimization on selceted region where i got all the models simlar to crystal structure with less difference between dope score(range ~5k to 6K)

Now, I want to what is the difference between automodel and loopmodel optimization on selected region ? and what loop model optimization exactly does ? and why there is a illogical contrast between DOPE score in loop model optimization?

Thanking you in advance

With Regards Mahesh V