Dear Modeller’s users,

I'm a real beginner with modeller (on Windows) and cannot manage to fix what seems to be a very simple case (according to what I've seen in the users' list).

I created my alignment file (1fgk.ali.txt)

>P1;pto-ad

sequence:pto-ad::::::::

MGSKYSKATNSINDALSSSYLVPFESYRVPLVDLEEATNNFDHKFLIGHGVFGKVYKGVL

RDGAKVALKRRTPESSQGIEEFETEIETLSFC-RHPHLVSLIGFCDERNEMILIYKYMEN

GNLKRHLYG------------SDLPTMSMSWEQRLEICIGAARGLHYLHTRAIIHRDVKS

INILLDENFVPKITDFGISKKGTELDQTHLSTVVKGTLGYIDPEYFIKGRLTEKSDVYSF

GVVLFEVLCARSAIVQSLPREMVNLAEWAVESHNNGQLEQIVDPNLADKIRPESLRKFGD

TAVKCLALSSEDRPSMGDVLWKLEYALRLQESVI*

>P1;1FGK_A

structureX:1fgk.ent:FIRST:A:LAST:A::::

------------ELPEDPRWELPRD---RLVLGKPLG-----QVVLAEAIGL----PNRV

TKVAVKMLKSDATEKD--LSDLISEMEMMKMIGKHKNIINLLGACTQDGPLYVIVEYASK

GNLREYLQARRPP------------EEQLSSKDLVSCAYQVARGMEYLASKKCIHRDLAA

RNVLVTEDNVMKIADFGLARDIHHIDYYKKTTNGRLPVKWMAPEALFDRIYTHQSDVWSF

GVLLWEIFTLGGSPYPGVP--VEELFKLLKEGHRMDKPSNCTN-ELYMMMR---------

---DCWHAVPSQRPTFKQLVEDLDRIVALTS---*

I also tried the following lines

structureX:1fgk.ent:464:A:762:A::::

structureX:1fgk:FIRST:A:LAST:A::::

Downloaded the corresponding pdb file (1fgk.ent)

And wrote the following script:

INCLUDE

SET ATOM_FILES_DIRECTORY = './:../../atom_files'

SET STOP_ON_ERROR = 2

SET OUTPUT_CONTROL = 1 1 1 1 2

#READ_MODEL FILE = '1fgk.ent', MODEL_SEGMENT '464:A' '762:A'

READ_MODEL FILE = '1fgk.ent', MODEL_SEGMENT 'FIRST:A' 'LAST:A'

SET ALNFILE='1fgk.ali.txt'

SET KNOWNS='1FGK_A'

SET SEQUENCE='pto-ad'

SET STARTING_MODEL=1

SET ENDING_MODEL=1

CALL ROUTINE = 'model'

I don’t have any “log” file as described in the manual but could redirect the comments into an error file which always give me the same message:

rdpdb___303E> No atoms were read from the specified input PDB file:

1) Possibly because an incorrect/non-existent PDB file is specified.

2) Possibly because the segment is specified incorrectly in the

alignment file or by the TOP variable MODEL_SEGMENT. That is,

the beginning residue number and/or chain id in MODEL_SEGMENT

may not be found in the input PDB file; MODEL_SEGMENT: 464: A

To find out more, switch on maximal output by 'SET OUTPUT_CONTROL = 1 1 1 1 2'

…………

pick_re_449W> No atoms selected; no restraints will be selected.

pick_re_450E> Probably because no MODEL is in memory; use READ_MODEL.

csrrng__299E> Atom index is out of bounds: 3 0

nselat__596E> Atom index is out of range (iatm,natm): 3 0

nselat__596E> Atom index is out of range (iatm,natm): 2 0

………….

At first, I got a message because the number of residues was different in the alignment file and the pdb file. I fixed it but it showed me that Modeller could access the pdb file.

I checked the residues’ numbers in the both files.

I tried to rename my file into 1fgk.pdb, to copy the pdb file in the modeller’s folder, to model only contiguous residues… and I have no more ideas.

All the files (alignment, PDB, top) are in the same folder and I could run another top script creating the alignment, using the data of the pdb file (what proves that it could find and read it…) (that is another problem but I was surprised to see that, if some residues were missing in the pdb file (in my case res 486 to 490) , Modeller considered the residues 485 and 491 to be contiguous. The alignment could then not be very accurate.)

Has someone any idea of what could get wrong in all of this?

* Modeller returns many files (V99990001, B99990001, D00000001 which are empty) and RSR file (full of data) and “Configuration settings “ file which gives me some very strange 3D results. What are all these files ?

* last question: is there any place where I could find the meaning of all the results, comments, warning and errors Modeller returns?

Many thanks for your help!

Muriel