Installation (Linux – debian based)
Go to this page : https://salilab.org/modeller/download_installation.html and download modeller for your system (32 or 64 bits).
Then execute this command :
sudo env KEY_MODELLER=XXXX dpkg -i modeller_9.14-1_i386.deb
with XXXX the modeller keys. You can ask the key through the registration form (https://salilab.org/modeller/registration.html). You will need a academic mail to get the key for free.
Usage
To use Modeller, you need 3 files : a alignement file, the reference PDB file, and a python model script (usualy called model-default.py).
- Alignement file : example (Color code : name of pdb file | first amino acid number (on pdb file) | Chain | End of the sequence | a name )>P1;3U5I
structureX:3U5I_D: 4:D:297: :RPL5 :: :
—QKDAKSSAYSSRFQTPFRRRREGKTDYYQRKRLVTQHKAKYNTPKYRLVVRFTNKDIICQIISSTITGDVVLAAAYSHELPRYGITHGLTNWAAAYA/ TGLLIARRTLQKLGLDETYKGVEEVEGEYELTEAVEDGPRPFKVFLDIGLQRTTTGARVFGALKGASDGGLYVPHSENRFPGWDFETEEIDPELLRSYIF/ GGHVSQYMEELADDDEERFSELFKGYLADDIDADSLEDIYTSAHEAIRADPAFKPTEKKFTKEQYAAESKKYRQTKLSKEERAARVAAKIAALAGQQ— —*
>P1;RPL5_H
sequence:test: 1 : : 297 : :Unknown :: : MGFVKVVKNKAYFKRYQVKFRRRREGKTDYYARKRLVIQDKNKYNTPKYRMIVRVTNRDIICQIAYARIEGDMIVCAAYAHELPKYGVKVGLTNYAAAYC/
TGLLLARRLLNRFGMDKIYEGQVEVTGDEYNVESIDGQPSAFTCYLDASLARTTTGNKVFGALKGAVDGGLSIPHSTKRFPGYDSESKEFNAEVHRKHIM/
GQNVADYMRYLMEEDEDAYKKQFSQYIKNSVTPDMMEEMYKKAHAAIRENPVYEKKPKKEVKK——KRWNRPKMSLAQKKDRVAQKKASFLRAQERA AES* - model-default.py (example)
#!/usr/bin/env python # Homology modeling by the automodel class from modeller import * # Load standard Modeller classes from modeller.automodel import * # Load the automodel class log.verbose() # request verbose output env = environ() # create a new MODELLER environment to build this model in # directories for input atom files #env.io.atom_files_directory = ’./:../atom_files’ #env.io.atom_files_directory = '/home/../../../modeller' a = automodel(env, alnfile = 'aligment.ali', # alignment filename knowns = '3U51', # codes of the templates sequence = 'RPL5_H') # code of the target a.starting_model= 1 # index of the first model a.ending_model = 100 # index of the last model # (determines how many models to calculate) a.make() # do the actual homology modeling
Be careful, « knowns » is the name of the template (first line « >P1;3U5I« ) and « sequence » is the name of the target (« >P1;RPL5_H« )
Execution :
mod9.14 model-default.py
If you have this error : « ‘import site’ failed; use -v for traceback », it’s not a big deal. Modeller will work without issues.
Get the Best model :
You need to analyse the objective function value. Use this command :
grep 'OBJECTIVE' file.B9999000*.pdb > FunctionObj.data
then in FunctionObj.data locate the lower value. Example :
NB2.B99990001.pdb:REMARK 6 MODELLER OBJECTIVE FUNCTION: 4752.2593 NB2.B99990002.pdb:REMARK 6 MODELLER OBJECTIVE FUNCTION: 4577.8940 NB2.B99990003.pdb:REMARK 6 MODELLER OBJECTIVE FUNCTION: 4632.4722 NB2.B99990004.pdb:REMARK 6 MODELLER OBJECTIVE FUNCTION: 4785.2485 NB2.B99990005.pdb:REMARK 6 MODELLER OBJECTIVE FUNCTION: 4628.4385 NB2.B99990006.pdb:REMARK 6 MODELLER OBJECTIVE FUNCTION: 4680.4165 NB2.B99990007.pdb:REMARK 6 MODELLER OBJECTIVE FUNCTION: 4538.2246 <--- NB2.B99990008.pdb:REMARK 6 MODELLER OBJECTIVE FUNCTION: 4616.3188 NB2.B99990009.pdb:REMARK 6 MODELLER OBJECTIVE FUNCTION: 4720.9233 NB2.B99990010.pdb:REMARK 6 MODELLER OBJECTIVE FUNCTION: 4669.3135
Here you best PDB is NB2.B99990007.pdb
Thank you very much your explaination was perfect