First, there is no need for minimization as Modeller already performs optimization . In general, use energy minimization for stereochemically incorrect models (usually not from Modeller) or as intro to molecular dynamics simulation (but don't use molecular dynamics to improve the model). Second, there is little sense in checking Ramachandran plot as 1) Modeller already generates stereochemically correct models, unless you did sth really bad with the alignment 2) phi and psi angles are based on the templates, so if the the template had a residue in disallowed region, the model will have it too.
I did not get your answer @Classes on linee........I am asking about the about 3D structure models of proteins..
am not sure...but modeller gives to you a list of scores for each model. You can choose the best model in function of their energy , lower energy , more stable is the model. Modeller gives two types of scores : DOPE score and GA341 score , for DOPE score you must choose the model with lower value (lower energy) and if you use Ga341 , the value should be close to 1. After you choose the best model, you can evaluate it , and plot the energies of the model via gnu plot ( linux) or in https://prosa.services.came.sbg.ac.at/prosa.php. Byee ! :) and sorry for my english.
its ok......but Modeller gives " Modeller lowest objective function score" present in PDB file of Model..How can i find/ Check DOPE or GA341 score...
mmm i use modeller in linux. Modeller gives me a file with the name model-single.LOG ( it contains info about the models) , at the end of this file you can see the scores , in this case dope score. I attached an example file.
how can i get this file by Modeller....?? what command and phython file is required for getting log file.
You should choose the one that is most similar to the template.
Due to optimization procedure the ten models can deviate from the template to different extent. This will indeed result in different energies, but relying on energies is not safe - unless DOPE score has been proven to significantly correlate with deviation to the native structure for ten closely similar models. Selecting the model that is most similar to the template is safer - you minimize a chance that your model have distortions due to optimization procedure.
First, there is no need for minimization as Modeller already performs optimization . In general, use energy minimization for stereochemically incorrect models (usually not from Modeller) or as intro to molecular dynamics simulation (but don't use molecular dynamics to improve the model). Second, there is little sense in checking Ramachandran plot as 1) Modeller already generates stereochemically correct models, unless you did sth really bad with the alignment 2) phi and psi angles are based on the templates, so if the the template had a residue in disallowed region, the model will have it too.
Dear Adnan,
Please the installation folder of modeller. Reaching this place, you will find a folder named 'examples'... or something like that, I exactly don't remember. Next you find the best example, which is actually a code to perform the modelling task. Chose your flavor, using that code, write a simple code.. or just modify it according to your input files. Alignment is the most important thing to remember; Even the Author Sali et al. have warned about this. prepare input files. And generate 100 models.
Chose the top ones using the assessment code 'Normalized Z dope' or something like that, which could be found in the installation directory.
The Z dope with value equal to -1 or less can be considered best models from those 100 generated. You can even get 20 models with value -1, or just 5 models or none. So have patience. And re try the same procedure using the another template.
And yes, please do not forget to cite Sali et al., as you are using their codes for modelling your protein.
- Badges
- Science topic
- Similar topics
- Bioinformatics
- Bioinformatics and Computational Biology