[Dock-fans] skipped compounds

Philip Crawford p.crawford at unsw.edu.au
Wed Mar 21 21:33:28 PST 2007


Scott,
I haven't run Dock 6.1 in MPI mode but Dock 6, when running in MPI mode,
only lists the compounds which get valid scores.

I was trying to run on a small binding site (benzoate was the inhibitor in
the known structure) and I found in non-MPI mode Dock6 was listing lots of
molecules with an error, suggesting the grid may not be large enough. This
doesn't happen when running a library through in MPI mode.

I can't recall but I think it reported the correct number of molecules
processed when running in MPI mode. Just didn't fill the Dock.out file with
messages about molecules which don't fit.

Cheers,

Phil Crawford

_____________________________________________
Philip Crawford B Comp Sc, MIEEE
Medicine Computer Support Unit
The University of NSW, Sydney, NSW, 2052
Phone: +61-2-9385 2564
Fax: +61-2-9385 1258
Email: p.crawford at unsw.edu.au
_____________________________________________

-----Original Message-----
From: dock-fans-bounces at docking.org [mailto:dock-fans-bounces at docking.org]
On Behalf Of Scott Brozell
Sent: Thursday, 22 March 2007 5:12 PM
To: Yuliet Mazola Reyes
Cc: dock-fans at docking.org
Subject: Re: [Dock-fans] skipped compounds

Hi,

On Wed, 21 Mar 2007, Yuliet Mazola Reyes wrote:

> I am docking a database of chemical compounds, and I noticed that some 
> compounds were not docked into the receptor and they are not reported 
> in the WARNING read_mol2 lines of the dock output file, which seems 
> that their MOL2 files shouldn't have problems to be read by Dock. I am 
> using the bump filter option to avoid bad orientations which overlap 
> with the receptor and do not score them.

My comments are based on the a brief look at the source code, but there are
a lot of paths through the code.  I am using this input
install/test/bump_filter/bump2.dockin
to verify my statements.

Every molecule in the ligand_atom_file should have its title emitted like
this, e.g.:

-----------------------------------
Molecule: ZINC00157467

And every molecule in the ligand_atom_file should be tallied in this line at
the tail of the dock output, e.g.:

12 Molecules Processed

Thus, you should be able to account for every molecule in the
ligand_atom_file.

> I would like to know if compounds that do not pass the bump filter are 
> counted or reported in "compounds skipped" line in the info file or 
> they will be just excluded by Dock, and are not reported in the info 
> file and either in a WARNING line in the output dock file?
> Is there an option to allow the user to know which compounds are 
> rejected by bump filter?

What is the "compounds skipped line in the info file" ?

It appears that there is no output what so ever when something is skipped
via the bump filter.
If docking of a molecule is not successful then either this:

ERROR:  Could not find valid anchor orientation.

or this:

ERROR:  Could not complete growth.

will be emitted.


> I also want to know if lines 13 and 14 which are blank in the original 
> SYBYL MOL2 file, separating  <TRIPOS>MOLECULE record from <TRIPOS>ATOM 
> record (see example of MOL2 file bellow), are ignored by Dock or they 
> should be kept blank to allow the reading of MOL2 file by Dock program.

If those lines are present then they are read and saved.
If those lines are not present then they are not read.
Thus DOCK is robust with respect to those optional data lines;
http://www.tripos.com/mol2/mol2_format25.html#1184

Scott

> SYBYL MOL2 EXAMPLE
>
> 1      #         Name: benzene
> 2      #         Creating user name: tom
> 3      #         Creation time: Wed Dec 28 00:18:30 1988
> 4
> 5      #         Modifying user name: tom
> 6      #         Modification time: Wed Dec 28 00:18:30 1988
> 7
> 8      @<TRIPOS>MOLECULE
> 9      benzene
> 10     12 12 1 0         0
> 11     SMALL
> 12     NO_CHARGES
> 13
> 14
> 15     @<TRIPOS>ATOM
> 16     1       C1      1.207   2.091   0.000  C.ar    1      BENZENE 0.000
> 17     2       C2      2.414   1.394   0.000  C.ar    1      BENZENE 0.000
> 18     3       C3      2.414   0.000   0.000  C.ar    1      BENZENE 0.000
> 19     4       C4      1.207   -0.697 0.000   C.ar    1      BENZENE 0.000
> 20     5       C5      0.000   0.000   0.000  C.ar    1      BENZENE 0.000
> 21     6       C6      0.000   1.394   0.000  C.ar    1      BENZENE 0.000
> 22     7       H1      1.207   3.175   0.000  H       1      BENZENE 0.000
> 23     8       H2      3.353   1.936   0.000  H       1      BENZENE 0.000
> 24     9       H3      3.353   -0.542 0.000   H       1      BENZENE 0.000
> 25     10      H4      1.207   -1.781 0.000   H       1      BENZENE 0.000
> 26     11      H5      -0.939 -0.542 0.000    H       1      BENZENE 0.000
> 27     12      H6      -0.939 1.936    0.000  H       1      BENZENE 0.000
> 28     @<TRIPOS>BOND
> 29     1       1       2       ar
> 30     2       1       6       ar
> 31     3       2       3       ar
> 32     4       3       4       ar
> 33     5       4       5       ar
> 34     6       5       6       ar
> 35     7       1       7       1
> 36     8       2       8       1
> 37     9       3       9       1
> 38     10      4       10      1
> 39     11      5       11      1
> 40     12      6       12      1
> 41 @<TRIPOS>SUBSTRUCTURE
> 42     1       BENZENE 1       PERM    0      ****    ****   0       ROOT
_______________________________________________
Dock-fans mailing list
Dock-fans at docking.org
http://blur.compbio.ucsf.edu/mailman/listinfo/dock-fans




More information about the Dock-fans mailing list