Computational approaches for investigating protein-ligand interactions - towards an in-depth understanding of the dengue virus methyltransferase

Schmidt, Tobias Benjamin. Computational approaches for investigating protein-ligand interactions - towards an in-depth understanding of the dengue virus methyltransferase. 2013, Doctoral Thesis, University of Basel, Faculty of Science.


Official URL: http://edoc.unibas.ch/diss/DissB_10596

Downloads: Statistics Overview


Interactions between proteins and their ligands play crucial roles in many
biological processes, such as metabolism, signaling, transport, regulation or
molecular recognition. Understanding the molecular basis of protein-ligand
interactions is thus of great interest, not only for modeling complex biological
systems but also for applications in drug discovery. However, structural details
for most of these interactions have not been characterized experimentally.
Therefore, computational methods have become increasingly important for
investigating biological systems at an atomistic level.
This work aims at a better understanding of the molecular basis of disease
related viral methyltransferases, their interactions with small molecules and
the catalytic mechanism, which may on the long perspective help to develop a
treatment against neglected tropical diseases. Furthermore, we aim to advance
the current methods for the computational prediction of a protein's molecular
function and its biological role in the cell. In addition, we aim to complement
currently available computational strategies for estimating protein ligand
interaction energies.
Dengue fever is a rapidly emerging, still neglected tropical disease which
causes significant mortality and morbidity in humans. For the discovery of novel
classes of compounds inhibiting dengue virus methyltransferase, a combination of
structure-based virtual screening and enzymatic inhibition assays is employed.
From the shortlist of 263 candidates selected by virtual screening, ten
compounds are found to specifically inhibit the target enzyme with IC50
values in the low uM range. Promising compounds are selected for further
experimental characterization and the inhibitory activity of the two most active
compounds is confirmed.
For obtaining a better understanding of the molecular basis of the target
enzyme's function, molecular dynamics simulations and mixed quantum
mechanics/molecular mechanics calculations are employed to investigate the
mechanisms of the enzymatically catalyzed reaction at an atomistic level. Based
on a structural model of the target protein in complex with its RNA substrate,
the impact of mutations on ligand binding, geometric arrangements and reaction
energy barriers are evaluated computationally. In addition, for a detailed
characterization of the underlying chemical reactions, ab initio electronic
structure calculations are performed on model systems approximating the
biological structure.
The reliable prediction of ligand binding sites is crucial for characterizing
proteins with unknown function. Therefore, the use of computational predictions
of protein function and ligand binding sites for proteins without experimental
structures are assessed in a blind and objective way. Limitations in the current
prediction methods are analyzed and suggestions for a more reliable evaluation
are given. Following those suggestions, an extended and fully automated
assessment is implemented in the Continuous Automated Model EvaluatiOn (CAMEO)
Computational identification of protein-ligand interactions can greatly
facilitate the drug discovery process. Thus, we establish a straightforward,
rapid scoring function that aims to identify the best poses out of an ensemble
of pre-docked poses, by quantifying the degree of burial and the electrostatic
interactions of the ligand in a binding site. The scoring function is evaluated
on a set of high quality protein-ligand complex structures, where the results
show promisingly high retrieval rates for selecting the best poses from a pool
of decoy poses.
Finally, a novel human-computer interface device is described which facilitates
the interaction with the computational representation of complex biological
systems by employing natural and intuitive movements.
Advisors:Schwede, Torsten
Committee Members:Meuwly, Markus
Faculties and Departments:05 Faculty of Science > Departement Biozentrum > Computational & Systems Biology > Bioinformatics (Schwede)
UniBasel Contributors:Schwede, Torsten and Meuwly, Markus
Item Type:Thesis
Thesis Subtype:Doctoral Thesis
Thesis no:10596
Thesis status:Complete
Number of Pages:189 S.
Identification Number:
edoc DOI:
Last Modified:22 Jan 2018 15:51
Deposited On:27 Jan 2014 14:07

Repository Staff Only: item control page