Gallery of SPROUT output screen images
A 24 bit TrueColor TIFF high resolution version of each image is provided
by clicking on the small inline images. The size of the large resolution
images vary from 0.1MB to 3MB.
APPA (cylinder bonds, carbon atoms are gold) in the binding site of Trypsin
(stick & ball representation, carbon atoms are green). The accessible
surface of the receptor is coloured according to the desired ligand interaction:
hydrogen bond acceptor ligand atoms are expected at red, donors at blue
patches. Yellow clouds represent hydrophobic pockets. Some hydrogen bonding
regions that are not satisfied by the ligand are indicated by solid objects:
the red is a place for H-bond acceptor atom, the light blue (with white
core) is for H-bond donor, the dark blue is for double donor and the magenta
is for dual group.
Guanine diphosphate (GDP: cylinder bonds, carbon atoms are gold) in the
binding site of P21 (stick & ball representation, carbon atoms are
green). The accessible surface of the receptor is rendered translucent
indicating the distance from the ligand atoms, i.e. close van der Waals
contacts are opaque and the translucency increases by the square of distance
between the van der Waals spheres until it becomes completely transparent
at 2.5A. Surface colouring and the solid objects have the same meaning
as on the previous image.
A different view of the GDP - P21 interaction site. In this image the complete
transparency is reached at 1.8A. A different set of H-bonding sites is
shown. The yellow patches of the surface indicate a combination of hydrophobic
and hydrogen bond acceptor activity requirement (red plus green gives yellow).
PPACK (cylinder bonds, carbon atoms are gold) in the active site of Thrombin
(stick & ball representation, carbon atoms are green). Accessible surface
and hydrogen bonding site representation as on the previous images.
The active site of the protein P21. Some potential interaction sites are
shown including the metal ion interaction which is indicated by a grey
region near the magenta magnesium ion. The hydrogen bonding regions are
shown with the same colour coding as on the previous images. The hydrophobic
regions are indicated by small yellow crosses.
A complete inventory of the interactions between GDP and P21. All the interaction
sites are shown that are satisfied by the ligand. Solid regions represent
interactions that are satisfied with geometric parameters within the default
tolerances of SPROUT. Other interactions (rendered by grid wire) occur
with parameters that require adjustment to the default tolerances, i.e.
these are weaker than average interactions. A key hydrophobic region is
rendered by translucent yellow surface allowing to see the guanine aromatic
system inside the hydrophobic pocket.
The accessible surface of the active site of P21 is shown in grid wire
representation with the bound ligand, GDP inside.
Small functional groups are docked by SPROUT to some of the key interaction
sites of P21 as a first step of generating novel potential ligands.
P21-GDP binding. The solvent accessible surface coloured by interaction
type (red acceptor, blue - donor, green - hydrophobic) is shown first without
the bound ligand in the upper half and then for the complex including the
ligand. The solvation effects of the ligand binding can be observed by
comparing the two pictures. The above one has a deep hydrophobic pocket
(green on the right side) which is filled by the ligand and the complex
shows mainly hydrophilic surface except for the sides of the pocket. Perhaps
a larger hydrophobic ligand could do a better job?
Interaction possibilities of P21 that are missed by GDP. This figure highlights
the missing interactions by clipping out the parts of the surface where
GDP can satisfy the needs of the protein. Another small hydrophobic pocket
(behind but too far from the sugar ring) emerges from this view suggesting
possible targets for a blocking drug.
Thermolysin complexed with a small inhibitor. The accessible surface is
coloured by the usual way (red acceptor, blue - donor, green - hydrophobic)
and Zinc binding site is shown in grey grid-wire. The hydrophobic sidechain,
the amide and the nitrogen of the phosphonamide have perfect interactions
but one of the oxygens of the phosphonamide seems to reside in a place
where hydrogen bond donor would be expected. The source of the surface
color is a glutamate residue which is assumed to be deprotonated, hence
electronegative. Is it possible that there is a protonated carboxylic acid
next to a Zn2+ ion ?
Two views (180 degree rotation about the vertical axes) of a larger inhibitor
complexed with thermolysin. More hydrophobic moieties, all of them placed
perfectly nearby hydrophobic (green) surface patches. It looks like a better
fit, indeed the biological activity of this ligand is higher by two order
of magnitudes than the previous.
Yet another thermolysin inhibitor in two views. The phenyl ring on the
left is a bit too far, it would be better above the hydrophobic area instead
of the oxygen atoms.
This color coded interaction surface offers an insight into the inhibition
of thrombin by TAPAP. The two views differ by a 90 degree rotation about
the vertical axes. The left view shows the hydrophobic fits at the open
part of the binding pocket and some hydrogen bonding from the sulfunamide
and the carbonyl behind. The right view shows the interaction of the benzamidine
with the deep pocket. The empty tunnel from this pocket with the hydrophobic
end offers an interesting opportunity to improve binding.
Two other thrombin inhibitor are compared by a paralell (compatible orientation)
view: argatroban (left) and NAPAP (right). It is appearent that both the
hydrophobic and the hydrogen bonding interactions are satisfied better
by NAPAP and the difference in biological activities supports this prediction.
Finally, the strongest thrombin inhibitor, PPACK is shown in two different
views. In fact, PPACK forms an irreversible covalent interaction with thrombin.
The covalent interaction possibility is predicted by SPROUT: there is a
small green region in the upper right corner of the right image. The oxygen
of the carbonyl group is involved in hydrogen bonding making the attached
carbon open for a nucleophilic attack by the oxygen of the serine residue.
The hydroxyl group is activated by deprotonation from the aspartate histidine
catalytic hydrogen bonding pattern (visible in front of the ligand in stick
and ball representation).
Growing a novel partial structure (cyan colour) from the phosphate binding
pocket of P21. The structure is anchored by the metal ion and complex hydrogen
bonding interaction sites. The growth is aimed towards the green sphere
representing a hydrophobic region.
The counterpart of the previous structure grown from the hydrogen bonding
group on the right towards the phosphate pocket on the left through the
hydrophobic region (satisfied by the phenyl ring).
The final solution set is clustered according to 2D similarity for the
sake of easier visual inspection. This picture shows 9 cluster centroids
and part of the dendogram representing the result of the hierarchical clustering.
The generated structures are sorted according to different complexity measures
(e.g. number of fusions, spiro joins, gauche interactions...). The picture
shows the 3 simpliest (top row) and the 3 most complicated structure out
of the 80 total generated by a half an hour run.