[Dock-fans] some question of rigid socre and time expened.

Trent E. Balius tbalius at aol.com
Tue Jul 23 06:40:54 PDT 2013


Dear LiLei,

I am also sending this to the dock-fans mailing list.

I understand that you are performing docking with rigid ligand:	

    flexible_ligand                                              no

This is a large approximation, the starting geometry of the ligands are 
likely not the geometries that will fit well in the pocket.  I believe 
that if you allow for ligand flexibility that this will help your 
enrichment:
    flexible_ligand                                              yes
    min_anchor_size                                              5
    pruning_use_clustering                                       yes
    pruning_max_orients                                          100
    pruning_clustering_cutoff                                    100
    pruning_conformer_score_cutoff                               100
    use_clash_overlap                                            no
    print_growth_tree                                            no

There are several parameters that you may wish to adjust to get better 
performance.

These changes might effect accuracy.   I would suggest changing the 
following two parameters:

    max_orientations                                             1000
to:
    max_orientations                                             100

    simplex_max_iterations                                       1000
to:
    simplex_max_iterations                                       100

See the following posting for more discussion on speed.

http://mailman.docking.org/pipermail/dock-fans/2012-April/002650.html

I hope this helps,

Trent



---- Original Message ----
From: LiLei <2012222040093 at stu.scu.edu.cn>
To: tbalius <tbalius at aol.com>
Sent: Sun, Jul 21, 2013 11:33 am
Subject: some question of rigid socre and time expened.


Dear Trent,
I,m a postgraduate from china and want to use dock6.6 for screening 
some inhibitors of our intereast target protein and I'm so sorry to 
trouble you again.

Previously, I have made some enrichment tests with rigid method.After 
my enrichment tests finished,almost all the AUC% were lower than that 
from DUD-E website.
So I want to know if other score function is better than rigid score 
for this purpose?

And i find that when use rigid method docked some ligand into target 
protien,the total elapsed time was 248 seconds ,some even more than 300 
seconds. My ligand library were more than 120,000 ligands so the time 
expended will very large.
some information of  dock.out as follows:


	Initializing Library File Routines...
	Initializing Orienting Routines...
	Read in 54 spheres for orienting.
	Initializing Grid Score Routines...
	 Reading the energy grid from grid.nrg
	 Done reading the energy grid.
	 Reading the grid box quantifiers from grid.bmp
	size = 839800
	spacing = 0.3
	origin = -5.795,-27.082,6.1185
	span = 95,85,104
	 Done reading the grid box quantifiers.
	 
	-----------------------------------
	VERBOSE MOLECULE STATS
	 
	  Number of heavy atoms = 29
	  Number of rotatable bonds = 5
	  Formal Charge = 0.00
	  Molecular Weight = 407.47
	  Heavy Atoms = 29
	-----------------------------------
	VERBOSE ORIENTING STATS :
	 
	Orienting 29 anchor heavy atom centers
	Sphere Center Matching Parameters:
	   tolerance: 0.25; dist_min: 2; min_nodes: 3; max_nodes: 10
	Num of cliques generated: 1622
	 Residual Info:
	   min residual:    0.0108
	   median residual: 0.1992
	   max residual:    0.2500
	   mean residual:   0.1902
	   std residual:    0.0464
	 Node Sizes:
	   min nodes:    3
	   max nodes:    4
	   mean nodes:   3.0012
	 
	-----------------------------------
	 
	 Elapsed time for docking: 248 seconds
	
	and my dock.in parameters as follows:
	
	
		ligand_atom_file                                             
./Ligand/13489.mol2
	
		limit_max_ligands                                            no
	
		skip_molecule                                                no
	
		read_mol_solvation                                           no
	
		calculate_rmsd                                               no
	
		use_database_filter                                          no
	
		orient_ligand                                                yes
	
		automated_matching                                           yes
	
		receptor_site_file                                           
selected_spheres.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                                              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                                       grid
	
		multigrid_score_secondary                                    no
	
		dock3.5_score_secondary                                      no
	
		continuous_score_secondary                                   no
	
		descriptor_score_secondary                                   no
	
		gbsa_zou_score_secondary                                     no
	
		gbsa_hawkins_score_secondary                                 no
	
		SASA_descriptor_score_secondary                              no
	
		amber_score_secondary                                        no
	
		minimize_ligand                                              yes
	
		simplex_max_iterations                                       1000
	
		simplex_tors_premin_iterations                               0
	
		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_random_seed                                          0
	
		simplex_restraint_min                                        no
	
		atom_model                                                   all
	
		vdw_defn_file                                               
 ./vdw_AMBER_parm99.defn
	
		flex_defn_file                                               
./flex.defn
	
		flex_drive_file                                             
 ./flex_drive.tbl
	
		ligand_outfile_prefix                                        rigid
	
		write_orientations                                           no
	
		num_scored_conformers                                        1
	
		rank_ligands                                                 no
		
		
		My box size was 8 angstrom.
		So how to change the  parameters of dock.in and can increase 
efficiency,because i have  changed some parameters but can not get 
better.May i get your help when you have time.
		
		Sincerely,
		
		LiLei




More information about the Dock-fans mailing list