Pharmacophore Descriptors for Sustituents
General remarks
Pharmacophore descriptors can be derived on a per substituent basis
by making use of rooted topological pharmacophores. These are
topological pharmacophores
in which one agon (pharmacophoric unit), the root, is singled out.
It corresponds to the point of attachment of the substituents.
This leads on the one hand to new moments, on the other hand to
restrictions in the superposition of pharmacophores in similarity
calculations.
The rest of these remarks is identical to the remarks in tutorial
Descriptors for Pharmacophoric Properties
Pharmacophoric properties of molecules are described in Moloc
mainly by topological pharmacophores
(-> theory). From these,
two types of descriptors have been
derived:
moment- and similarity- ones.
This tutorial illustrates how to calculate them for the example
of a set of substituted hydroxyquinolone compounds for which
plasma protein binding data are known (J.Med.Chem.40,4053,1997).
Files, containing the structures, hdrqn.sd, and the
experimental data, hdrqn.lst, can be found in the
moloc/dat directory.
Generation of Topological Moment Descriptors
Because the generation of substituent descriptors requires a
beforehand definition of a carrying scaffold, this option is
implemented only in interactive mode.
- Start Moloc with the parameter hdrqn.sd,
Moloc hdrqn.sd.
- Go to the library submenu 'lib/d'
(definition, descriptors etc. of a scaffold library).
- Define the scaffold (fragment) on one particular structure
'f':
- Select the structure (.e.g. hdrqn_59). You are then thrown
into the set menu to define the scaffold atoms.
- Specify scaffold atoms, e.g. with loop option
'l', by picking atoms N14, C8, and Cl13 in
that order.
- Exit from the set menu: The program confirms with
'substructure defined'.
- Locate the scaffold in the other structures
(option 's'), the program confirms with:
New entry set: frgmch1
Found 59 structures containing the fragment
- By default, all structures participate in the following
actions. If you want to restrict those to a subset, do so with
option 'm', which presents a selector of the
structures containing the fragment.
The only positions at which the substituens vary are C2, C4,
and C8 (on structure hdrqn_59, which is favorably set as the
only visible one). One can either perform a full descriptor
generation by selecting all three positions, or just for a
single position, e.g. C8, which carries the most divers set
of substituents. The first option yields a full description
of the varying parts of the set of molecules. It has the
disadvantage of specifying all substituents at the same
level of description (order of the moments). Selecting each
substituent position separately offers the advantage of
adjustment of the description level to the substituent
diversity. This, at the cost of having to concatenate files
for a full description. We just treat position 'C8':
- Select 'a' (pick fragment atoms for residue
data) and pick atom 'C8'.
- Choose 'r' (write a file of residue-specific
descriptors) and specify a file name for the table.
An empty choice produces output to the dialog window.
- On the appearing data selector set the topological
pharmacophore option to the maximum value of three and exit.
- The program now calculates 68 descriptors per substituent.
On the dialog window the atom names of the specified
position(s) will be confirmed.
Topological Similarity Descriptors
Generation of Files with Substituent Descriptors
- Repeat items 1 to 5 of the previous paragraph.
- Select 'a' and pick atoms 'C2', 'C4',
and 'C8', the atoms for which the substituents vary.
- Now click 'q' (write a file of similarity descriptors
for picked atoms) and specify a file name.
- From the next selector (which appears only, when more than
one position was specified) take the default choice
'position dependent models'.
- The program now calculates the descriptors, informs you about
the number of descriptors it obtained for each position
(2 at 'C2", 1 at 'C4', and 10 at 'C8'), and writes the
descriptor table file.
- The number of descriptors can be influenced to some extend by
altering the value of the 'loading threshold'
in menu point 'p' (set parameters for topological
pharmacophores). Higher thresholds yield fewer
descriptors.
The resulting table can now be used to calculate a (linear) model
for any measured properties of the 59 compounds.
Moloc now throws you into a new menu 'subtpr'
Options in menu 'subtpr' regarding substituents
Option 'a' (define action) offers several
possibilities to examine or act on substituents at a given
position. For our example we illustrate two options:
analyse, and evaluate.
The latter, which includes ranking of substituents,
requires coefficients of a linear model.
The evaluate option only appears, when
these coefficients have been entered into Moloc with
option 'c' (read model coefficients).
Analyse the Similarity Matrix of a Substituent Set
- Click 'a' (define action) and select
analyse.
- Choose the substituent position for which the analysis
is required. In our case only substituent 3, 'C8' is of
interest.
- Select '3' (the position number) to
be thrown into the cluster analysis menu
'clan' which is explained in the tutorial
Diversity Analysis of Pharmacophoric
Properties.
Ranking of Substituents
Substituents can be ranked according to their effect in
a given linear model at a particular position.
- Select 'c' (read model coefficients)
and specify the coefficient file (consult the help of
option 'c' for the required format).
- Select 's' (separate model substituents)
to produce substituent libraries (see next section) for the
purpose of demonstration. More realistically,
you might want to provide your own substituent library.
- Entries are recognized as substituents, when they have a
single explicit hydrogen atom attached at the atom which
makes the bond to the scaffold (i.e. H symbolizes the
fragment-atom of attachment). Such substituent sets must be
generated beforehand and are best put into a library
(generated in 'lib/g, lib/n' or upon reading
a multi-entry file). Of course such libraries can still be
read in at that place with option './g'.
- Click 'a' (define action) and select
evaluate.
- Now select '3' (the position number)
and specify your substituents (preferably library-wise) to
obtain a sorted list of the these substituents together
with their influence on the model.
Separating Substituents
Separating substituents from the model compounds can be
achieved with menu option
's' (separate model substituents), (see
point 2 of the previous section). The chopped off (different)
substituents, new Moloc entries, are gathered into libraries
(entry-sets) called 'sub_1', 'sub_2', .... The point of
attachment is represented by an additional atom, the type of
which can be chosen in menu option
't' (atom type for attachment). By default
this is an explicit hydrogen, and this is the only choice
recognized by Moloc in substituent ranking.