|
WARNING: wrong left boundary detected.
This warning indicates, that your input structure has an unusual geometry (eg. rule based conformation generation). You should run an energy minimization on your structures.
[top ]
Can I change the individual scoring function terms?
The FlexS scoring function is derived from the Böhm empirical scoring function (H.J.Böhm, The development of a simple scoring function to estimate the binding constant for a protein-ligand complex of known three dimensional structure, JCAMD, 8 (1994), 243-256.).
There is a short description of the FlexS scoring function in the appendix of our the J.Med.Chem. paper (C.Lemmen, T.Lengauer and G.Klebe, FlexS: A Method for Fast Flexible Ligand Superposition, JMC, 41 (1998), 4502-4520.).
The individual coefficients of the scoring function are loaded from the geometry_sp.dat:
@global_energy_parameters_sp PHASE G_ED G_CHG G_HYD G_DON G_ACC G_DONG G_ACCG G_OVL G_STM G_ROT
[top ]
Can FlexS accept a template molecule which is really a collection of superimposed molecules?
This can be done with FlexS, but our solution is not working on multi-mol2 files for the following reason: Our superpositioning algorithm needs a connected molecule graph. At the moment we can't handle disconnected components.
But what you can already do (since FlexS v1.7), is to merge the gaussian representations of several pairwise alignments. So you get the molecule graph of one reference structure enriched with the gaussians of some alignments. You can download a script.
[top ]
How do I select a base fragment?
It is now possible to run FlexS fully automatically for virtual screening purposes. We have adapted FlexS to this task, so that it can be run in batch mode in order to process huge amounts of data. In contrast to the JMC paper, we have implemented a method to select the base fragment automatically:
FLEXS/TEST_LIG> selbas a f
If you want to generate a superpostion for some QSAR analysis, you can improve the FlexS method by careful choices of the 'correct' reference structure and base fragment, eg. you know some core structure of the ligands. So the degree of automation depends on the application.
[top ]
Why doesn't FlexS produce no solution at all in some cases?
There is a volume overlap threshold which prevents FlexS from generating superposition solutions, where the test-ligand extends to far away from the reference-ligand. You can influence this parameter with:
FLEXS> set VOLUME_OVERLAP_TH
[top ]
What kind of preparation do the input compounds need?
Both, the reference compound as well as the test compound need to have partial charges and hydrogen atoms annotated. The test compound should also be in a reasonable low-energy conformation since the bond lengths and -angles are taken directly from the input structure. Deviations from standard-angles may result in poor recognition of the involved fragments which is used in multiple places to annotate properties like Hydrogen-bonding capabilities and preferred torsion angles.
[top ]
What is RigFit? How does it relate to FlexS?
RigFit is a module within the FlexS-program. It allows rigid-body superpositioning of molecules and molecular fragments onto the reference molecule. The RigFit-method has been described in detail in
C. Lemmen, C. Hiller, and T. Lengauer
RigFit: A new approach to superimpose ligand molecules
Journal of Computer-Aided Molecular Design, 12 (1998), 491-502.
The RigFit approach is used in the base placement procedure (PLACEBAS) if either the fragment to be placed is not suitable for the triangle placement routine or the user determines it to do so. Also, by selecting SUPERPOSITION_MODE 1 or 2 the use of RigFit within FlexS is triggered. See the User Guide for details.
[top ]
|
