|
Structural and Chemical Biology @ Vanderbilt University |
Members: Gordon Lemmon
all research projects
Background
Small molecule or ligand docking software predicts the interaction between a protein and a small molecule. Such software can be used to predict how a putative drug will react with its target protein, or how a mutation in a protein might affect its interaction with a small molecule. Small molecule design involves constructing ligands and using docking techniques to predict how they will interact with their target proteins. Current docking and design software often fails to model the flexibility seen in proteins or ligands with many rotatable bonds. Also, while many complexes involve multiple ligands, cofactors, and ions, current software can consider only one ligand at a time.
We are adapting RosettaLigand docking code to overcome these shortcomings.
Results
In order to more efficiently model ligand flexibility, we split ligands into multiple fragments.
Then we search the cambridge structural database for conformations of these fragments that are
seen in nature (figure 1). RosettaLigand docking software was modified to sample from among all
the conformations of each fragment during the docking step (figure 2).
Figure 1: all conformer carbons
Acetylpepstatin was split into a small fragment (maroon) and a large fragment (cyan). 1001 conformers
were generated for the small fragment, 293 for the large fragment
Figure 2: docked conformer carbons
Acetylpepstatin fragment conformers shown in figure 1 were docked with HIV-1 protease. 300 models
were generated.
Figure 3: Bright conformer ensemble
We have modified RosettaLigand to allow multiple small molecules to be docked simultaneously.
Dethiobiotin Synthetase docked with 5 ligands. 100 models are shown. Mg ions: blue and red, phosphate:
yellow, ADP: green, dethiobiotin: green.
Future work
An application for ligand design within the Rosetta coding framework will utilize a fragment-extension
approach. This involves docking a starting fragment, and then sampling from a library of extender
fragments to find extensions that increase binding energy (figure 4).
Figure 4: Ligand design flowchart
Possible fragment extension algorithm for ligand design. Blue boxes represent design steps that have
not yet been developed.