[Dock-fans] Write multiple conformations during parallel processing

Cen Gao cengao at gmail.com
Wed Jul 21 08:24:09 PDT 2010


Dear Sudipto:

Thanks for the email and the sample input. After comparing with your input
with mine, I think I know my problem here. It turned out to be that I should
set "cluster_conformations = yes" otherwise dock6 will only write one
conformation per molecule. (I have tested this on a set of 5,000
molecules.)Not exactly sure why though, as I originally thought if I
set it to no, dock
will output the top N poses simply based on the primary score.


Cen

On Fri, Jul 16, 2010 at 1:04 PM, Sudipto Mukherjee <sudmukh at yahoo.com>wrote:

> Dear Cen,
>
> A maximum of 3 docked conformers will be written to prefix_scored.mol2
> Nothing will be written to prefix_confomer.mol2 with write_conformations
> set to no.
>
> In your case, what is happening is that after scoring, pruning and
> clustering, only one viable docked conformation remains. You will get up to
> 3 conformations in prefix_scored.mol2 if you keep trying with different
> ligands. I suggest that try this with a small database of 1000 ligands
> instead of just docking in one molecule. You should that many of them will
> have 2 or more conformers in the prefix_scored.mol2 file.
>
> For your reference, here is the dock input file I used to verify this:
>
> orient_ligand                                                yes
> automated_matching                                           yes
> receptor_site_file
> 121P.rec.clust.close.sph
> max_orientations                                             1000
> critical_points                                              no
> chemical_matching                                            no
> use_ligand_spheres                                           no
> use_internal_energy                                          yes
> internal_energy_rep_exp                                      12
> flexible_ligand                                              yes
> min_anchor_size                                              5
> pruning_use_clustering                                       yes
> pruning_max_orients                                          1000
> pruning_clustering_cutoff                                    100
> pruning_conformer_score_cutoff                               100.0
> use_clash_overlap                                            no
> print_growth_tree                                            no
> bump_filter                                                  no
> score_molecules                                              yes
> contact_score_primary                                        no
> contact_score_secondary                                      no
> grid_score_primary                                           yes
> grid_score_secondary                                         no
> grid_score_rep_rad_scale                                     1
> grid_score_vdw_scale                                         1
> grid_score_es_scale                                          1
> grid_score_grid_prefix                                       121P.rec
> dock3.5_score_secondary                                      no
> continuous_score_secondary                                   no
> gbsa_zou_score_secondary                                     no
> gbsa_hawkins_score_secondary                                 no
> amber_score_secondary                                        no
> minimize_ligand                                              yes
> minimize_anchor                                              yes
> minimize_flexible_growth                                     yes
> use_advanced_simplex_parameters                              no
> simplex_max_cycles                                           1
> simplex_score_converge                                       0.1
> simplex_cycle_converge                                       1.0
> simplex_trans_step                                           1.0
> simplex_rot_step                                             0.1
> simplex_tors_step                                            10.0
> simplex_anchor_max_iterations                                500
> simplex_grow_max_iterations                                  500
> simplex_grow_tors_premin_iterations                          0
> simplex_random_seed                                          0
> atom_model                                                   all
> vdw_defn_file
> vdw_AMBER_parm99.defn
> flex_defn_file                                               flex.defn
> flex_drive_file                                              flex_drive.tbl
> ligand_outfile_prefix                                        flex.dock2grid
> write_orientations                                           no
> num_scored_conformers                                        1
> write_conformations                                          no
> rank_ligands                                                 no
>
> Regards
> Sudipto Mukherjee
> Trent E. Balius
> Graduate Student, Robert C. Rizzo Lab
> Dept. of Applied Math & Statistics, Stony Brook University
>
> ------------------------------
> *From:* Cen Gao <cengao at gmail.com>
> *To:* Sudipto Mukherjee <sudmukh at yahoo.com>
> *Cc:* Dock-fans at mailman.docking.org
> *Sent:* Thu, July 15, 2010 6:03:42 PM
> *Subject:* Re: [Dock-fans] Write multiple conformations during parallel
> processing
>
> Hi Sudipto:
>
> Thanks for the email. I have tried to set  "num_scored_conformer" to 3 and
> "write_conformations" to No before, but it still generated only one pose per
> molecule to prefix_scored.mol2, while never wrote anything to
> prefix_confomer.mol2. May I ask whether this setting works differently on
> your version of DOCK?  (I am using DOCK6.4, compiled with GCC4.1)  Or are
> there any parameters I should set?
>
> If I set write_conformations to yes, it was able to write 1 pose to
> prefix_scored.mol2 and 3 poses per molecule into prefix_confomer.mol2 as
> expected. But again it works in serial version of DOCK only, but refuses to
> run in parallel.
>
>
> Thanks
>
>
> Cen
>
>
> On Thu, Jul 15, 2010 at 5:21 PM, Sudipto Mukherjee <sudmukh at yahoo.com>wrote:
>
>> Hi Cen,
>>
>> As you mentioned in your email, the "num_scored_conformer" is the
>> parameter you want to use.
>> Set this to 3 as you wish.
>> Set "write_conformations" as no. By default this is yes. The naming
>> convention is a bit confusing here, I admit. The write_conformations
>> parameter writes out partial conformers in a separate conformers.mol2
>> Setting this parameter to yes is what geenrates the ERROR about filling up
>> the disk during mpi runs.
>>
>> Now you should have a maximum of 3 conformers for each molecule screened.
>>
>> Regards
>> Sudipto Mukherjee
>> Graduate Student, Robert C. Rizzo Lab
>> Dept. of Applied Math & Statistics, Stony Brook University
>>
>> ------------------------------
>> *From:* Cen Gao <cengao at gmail.com>
>> *To:* Dock-fans at mailman.docking.org
>> *Sent:* Tue, July 13, 2010 12:13:53 PM
>> *Subject:* [Dock-fans] Write multiple conformations during parallel
>> processing
>>
>> Dear Dock fans:
>>
>> May I ask is it possible for DOCK to write out the top N conformers for
>> each molecule it processed during the parallel docking run? The manual
>> states that this feature is only available to serial version of DOCK to
>> "protect against filling up the disk", as "in conformer.mol2, the number of
>> molecules in the output file will be equal to the database size * the number
>> of anchors per molecule * the number of conformations that pass the final
>> pruning stage * the number of conformers per cycle."
>>
>> But if I understand it right, this conformer number should not be that
>> huge as it is something we can control using the keyword
>> "num_scored_conformer" (in the case of primary_score only), right?  As long
>> as I set this keyword to a smaller value, say 3 or 5, it will only write 3
>> or 5 poses per molecule, which will not be a problem in case of the file
>> size. Did I miss something here?
>>
>>
>> Thanks
>>
>> Cen
>>
>>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.docking.org/pipermail/dock-fans/attachments/20100721/fac51e88/attachment-0001.html 


More information about the Dock-fans mailing list