problem with BLK residues using multiple templates

[Lucy Forrest <>]

Hi all,

I've been testing out using Modeller to build ferredoxins (some of the 
ones in the examples, actually) containing 4Fe-4S cluster(s). I wanted 
to try different levels of description, so to start out I've been 
describing the FeS clusters as BLK residues. This works very well for 
single-template modeling, but when I try to use e.g. 3 template 
structures, I get 3 times as many atoms in the model cluster as I 
wanted. I've attached the top and alignment files but there are 3 input 
atom files (plus the model), so I'll send those on request (in the input 
atom files I use, the residue names have been changed to BLK and some of 
the atomnames have been changed from the original pdb, so that they all 
describe the same position: using the original pdbs should still give 
you the same problem of excess atoms, but the clusters will be distorted).

Does anyone have a reason for this, or a way out (apart from deleting 
the 'extra' atoms afterwards)?
I'll be building a new FS4 residue anyway; has anyone done that previously?

Many thanks in advance,
Lucy Forrest

# Modeller input script
# LRF 14mar02
# Build models of 2fdn using 3 template pdbs simultaneously
# and multiple alignment
#
# This script should produce one model, 2fdn.B99990016 
# try HETERO atoms
#

SYSTEM COMMAND = 'cp 2fdn_3_template_1.ali~ 2fdn_3_template_1.ali'

INCLUDE                             # Include the predefined TOP routines

SET OUTPUT_CONTROL  = 1 1 1 1 1

SET ALNFILE  = '2fdn_3_template_1.ali'# alignment filename
SET KNOWNS   = '1DUR' '1BLU' '1CLF' # codes of the templates
SET SEQUENCE = '2fdn'               # code of the target
SET HETATM_IO = 'on'                # use Fe/S heteroatoms
SET ATOM_FILES_DIRECTORY = '../atom_files.d/'    # directories for input pdb 
SET STARTING_MODEL= 16              # index of the first model 
SET ENDING_MODEL  = 16              # index of the last model
SET FINAL_MALIGN3D = '1'            # make fitted structures

# thorough MD only 
#SET MD_LEVEL = 'refine_4'

# Repeat the whole cycle x times and do not stop unless obj.func. > 1E6
SET REPEAT_OPTIMIZATION = 3, MAX_MOLPDF = 1E6

CALL ROUTINE = 'model'              # do homology modelling

>P1;1DUR
structureX:1DUR:1    :A:102  :A:::-1.00:-1.00
AYVINDSCIACGACKPECPVNCIQEG-SIYAIDADSCIDC------GSCASVCPVGA-----PNPED--------
-----/..*
>P1;1BLU
structureX:1BLU:1    : :102  : :::-1.00:-1.00
ALMITDECINCDVCEPECPNGAISQGDETYVIEPSLCTECVGHYETSQCVEVCPVDCIIKDPSHEETEDELRAKY
ERITG/..*
>P1;1CLF
structureN:1CLF:1    : :102  : :::-1.00:-1.00
AYKIADSCVSCGACASECPVNAISQGDSIFVIDADTCIDC------GNCANVCPVGAPVQE--------------
-----/..*
>P1;2fdn
sequence:2fdn:.    :.:.    : :::-1.00:-1.00
AYVINEACISCGACEPECPVNAISSGDDRYVIDADTCIDC------GACAGVCPVDAPVQA--------------
-----/..*