Difference between revisions of "Lab Journal of Task 5 (MSUD)"

From Bioinformatikpedia
(→‎Swissmodel)
m (→‎Multiple template modelling)
 
(13 intermediate revisions by 2 users not shown)
Line 1: Line 1:
 
== Calculation of models ==
  
== Calculation of models ==
  +
  +
Structure models were build with the following programs, using 2BFE and 3EXG as templates.
   
 
=== Modeller ===
  
=== Modeller ===
  +
We have written a Python script '''[[Modeller.py]]''' which is a wrapper for the original '''modeller''' program. It automatically calls '''modeller''' and generates models using single or multiple templates.
  +
  +
==== Single template modelling ====
  +
For single template modelling, we have used the structure 2BFE (99% sequence identity to 1U5B) and 3EXG (24.9% sequence identity to 1U5B).
  +
Computed models are stored in the directory <tt>/mnt/home/student/weish/master-practical-2013/task05/models</tt> on the student server.
  +
  +
==== Multiple template modelling ====
  +
In order to evaluate the performance of homology modelling, we have also tried to use multiple structures with homologous sequence as templates for structure prediction. The Python program <tt>create_dataset.py</tt> ([[Task_2_lab_journal_(MSUD)#Dataset_creation]]) was used for searching for homologous PDB structures with high and low sequence identities.
   
 
=== Swissmodel ===
  
=== Swissmodel ===
   
−
Swissmodel was run on [http://swissmodel.expasy.org/workspace/index.php?func=modelling_simple1&userid=USERID&token=TOKEN this server].
 +
Swissmodel was run on [http://swissmodel.expasy.org/workspace/index.php?func=modelling_simple1&userid=USERID&token=TOKEN this server]. The models are stored in <code>/mnt/home/student/schillerl/MasterPractical/task5/Swissmodel/</code>.
   
 
=== iTasser ===
  
=== iTasser ===
  +
  +
[http://zhanglab.ccmb.med.umich.edu/I-TASSER/ This server] was used for running iTasser. For creating a model with the low sequence identity template (3EXG), "Option II" was set to exclude templates with more than 30 % sequence identity from the iTasser template library. Models are stored in <code>/mnt/home/student/schillerl/MasterPractical/task5/iTasser/</code>.
   
 
== Evaluation of models ==
  
== Evaluation of models ==
  +
  +
The calculated models were compared to the reference structure 1U5B.
  +
Since the residue numbering in the 1U5B pdb file did not fit the query protein sequence of BCKDHA, we used the following script to change the numbering in the pdb file.
  +
  +
  +
<source lang="python">
  +
'''
  +
Change residue numbering in pdb file by adding an offset.
  +
  +
Usage: python change_residue_numbering.py <pdb file> <chain> <offset> <output file>
  +
  +
@author: Laura Schiller
  +
'''
  +
  +
import sys
  +
  +
pdb_file = open(sys.argv[1])
  +
out_file = open(sys.argv[4], "w")
  +
chain = sys.argv[2]
  +
offset = int(sys.argv[3])
  +
  +
line = pdb_file.readline()
  +
while(line):
  +
if (line.startswith("ATOM") or line.startswith("TER")) and line[21:22] == chain:
  +
number = int(line[23:26]) + offset
  +
new_line = "%s%3d%s" % (line[0:23], number, line[26:len(line)])
  +
out_file.writelines(new_line)
  +
else:
  +
out_file.writelines(line)
  +
line = pdb_file.readline()
  +
  +
pdb_file.close()
  +
out_file.close()
  +
</source>
  +
   
 
=== GDT ===
  
=== GDT ===
  +
  +
For calculating GDT_TS values, the [http://proteinmodel.org/AS2TS/LGA/lga.html LGA] server was used with options <code>-3 -sda -ch1:A -ch2:A</code>.
   
 
=== RMSD ===
  
=== RMSD ===
  +
  +
RMSD values were calculated on the [http://mathbio.nimr.mrc.ac.uk/wiki/SAP SAP server].

Latest revision as of 13:07, 2 August 2013

Calculation of models

Structure models were build with the following programs, using 2BFE and 3EXG as templates.

Modeller

We have written a Python script Modeller.py which is a wrapper for the original modeller program. It automatically calls modeller and generates models using single or multiple templates.

Single template modelling

For single template modelling, we have used the structure 2BFE (99% sequence identity to 1U5B) and 3EXG (24.9% sequence identity to 1U5B). Computed models are stored in the directory /mnt/home/student/weish/master-practical-2013/task05/models on the student server.

Multiple template modelling

In order to evaluate the performance of homology modelling, we have also tried to use multiple structures with homologous sequence as templates for structure prediction. The Python program create_dataset.py (Task_2_lab_journal_(MSUD)#Dataset_creation) was used for searching for homologous PDB structures with high and low sequence identities.

Swissmodel

Swissmodel was run on this server. The models are stored in /mnt/home/student/schillerl/MasterPractical/task5/Swissmodel/.

iTasser

This server was used for running iTasser. For creating a model with the low sequence identity template (3EXG), "Option II" was set to exclude templates with more than 30 % sequence identity from the iTasser template library. Models are stored in /mnt/home/student/schillerl/MasterPractical/task5/iTasser/.

Evaluation of models

The calculated models were compared to the reference structure 1U5B. Since the residue numbering in the 1U5B pdb file did not fit the query protein sequence of BCKDHA, we used the following script to change the numbering in the pdb file.


<source lang="python"> Change residue numbering in pdb file by adding an offset.

Usage: python change_residue_numbering.py <pdb file> <chain> <offset> <output file>

@author: Laura Schiller

import sys

pdb_file = open(sys.argv[1]) out_file = open(sys.argv[4], "w") chain = sys.argv[2] offset = int(sys.argv[3])

line = pdb_file.readline() while(line): 

   if (line.startswith("ATOM") or line.startswith("TER")) and line[21:22] == chain:
       number = int(line[23:26]) + offset
       new_line = "%s%3d%s" % (line[0:23], number, line[26:len(line)])
       out_file.writelines(new_line)
   else:
       out_file.writelines(line)
   line = pdb_file.readline()

pdb_file.close() out_file.close() </source>


GDT

For calculating GDT_TS values, the LGA server was used with options -3 -sda -ch1:A -ch2:A.

RMSD

RMSD values were calculated on the SAP server.

Retrieved from "https://i12r-studfilesrv.informatik.tu-muenchen.de/wiki/index.php?title=Lab_Journal_of_Task_5_(MSUD)&oldid=33883"