Structure-based Drug Design
Structurally-informed
approaches have increasingly demonstrated their value in drug design since the
first biologically-relevant X-ray structures became available 30 years ago. The
impact of these methods and technologies on early lead discovery and lead
optimization is significant. Issues that are of current relevance include:
- Are we maximizing the use
of (the never-ending, increasing) current computer power in Structure-based
Drug Design (SBDD)?
- Virtual Screening (VS) is usually applied to enrich datasets with
high-activity compounds. The "unusual" application of VS to weaker
kinase binders is an interesting area of exploration.
- Cross-docking applied to a structurally-rich CDK2 dataset can shed some light
on the pros and cons of utilizing docking methods during lead optimization.
- What is the function of modeling water molecules in SBDD? Instead of ignoring
(or deleting) them, their influence on binding affinity should be considered.
- What do we know, what do we we think we know or simply don’t know about SBDD?
On 17 October 2007 we will hold an eCheminfo Community of Practice conference session at Bryn Mawr College, Philadelphia to discuss latest advances in SBDD. The session will be chaired by Jose Duca (Schering-Plough) and includes a knowledgeable panel of speakers and discussion leaders: Daniel Cheney (Bristol-Myers Squibb), Natasja Brooijmans (Wyeth), Jose Duca (Schering-Plough), Terry Stouch (JCAMD) and Julian Tirado-Rives (Yale). A description of the session with presentation abstracts follows:
Structure-based Drug Design
http://echeminfo.com/COMTY_conferencesprog07sbdd
(Please
follow continuation to read abstracts)
Abstracts
Taking
advantage of current computational capacities: Applications of high-resolution
techniques in computer-assisted drug design
Daniel L.
Cheney, Department of Molecular Biosciences, Bristol-Myers Squibb, Hopewell, New Jersey
Current computational resources have grown enormously over recent years, to the
extent that modern gaming PC’s easily surpass on a per-processor basis the
performance of supercomputers of a decade ago. The emergence of Linux clusters,
office-PC grids, and cheap storage and memory is challenging the modeling
community to integrate into the drug design process higher resolution
methodologies which were until recently beyond our resources. In this
presentation, the utilization of relatively rigorous techniques in the context
of routine molecular modeling will be discussed. Applications involving real
world problems in drug discovery will also be described
Realistic
Virtual Screening Assessment in Kinases
Natasja
Brooijmans, Wyeth
Most retrospective receptor-based virtual screening studies in the literature
enrich libraries with highly-active compounds against the studied targets. While
these studies show that docking can be used successfully to distinguish active
ligands among decoys, these are not very realistic. In industrial settings,
corporate and commercial collections are screened that contain very few
optimized compounds that inhibit the target of interest in the nanomolar range.
Leads identified by high-throughput screening and virtual screening are
generally in the low-to-medium micromolar range. Using Glide SP docking, we
have performed studies on a set of kinase targets that show that identifying
these weak leads is significantly more difficult than identifying
highly-optimized ligands. Enrichment Factors obtained with Glide4.0 and 4.5
will be compared. Finally, the use of pre- and post-docking filters to enhance
Enrichment Factors was investigated.
Cross-docking
vs. Scoring: Is Overfitting the Third Wheel?
José Duca,
Schering-Plough Research Institute
Addressing protein flexibility has been a shortcoming of many methods. In this
study, we used cross-docking of ~150 inhibitors into the full set of crystal
structures for each inhibitor complexed with the kinase CDK2. In scoring
relative binding potency based on multiple combinations of several target
proteins, the dangers of over-fitting became apparent. Examples will be given
of insights gained into ligand properties such as pKa values and relative
tautomeric stabilities computed via ab initio quantum mechanical methods.
Ligand/protein
binding in Structure-Based Drug Design: Examples of the role of water and
caveats in its treatment
Terry R. Stouch,
Editor-in-Chief of the Journal of Computer-Aided Molecular Design
The importance of the role of solvation in drug binding is often alluded to and
explicit roles of individual water molecules have been recognized in some
drug-protein complexes. However, the energetics and extent of water's
importance has not been clearly defined and water molecules are not always seen
by experiment and can be mistaken for other things. We will present clear
examples of complexes where water plays active and passives roles in
drug-protein binding and sometimes a major role. Computational analysis will be
described that helps to understand the extent, impact, and energetics of
water's effects with comparison to some more commonly used methods to describe
solvation.
Ideas,
Approaches and Progress in Structure-Based Drug Design
Julian
Tirado-Rives, Yale University, Department of
Chemistry
The process of going from a known structure of a pharmaceutical target to an
active compound may seem both conceptually simple and practically daunting. This
talk will expand on our current thoughts and ideas on the overall process, the
computational approaches we utilize and are currently exploring, some of the
progress achieved and future directions. Emphasis will be placed on the
development of non-nucleoside inhibitors for HIV-1 reverse transcriptase using
Free Energy Perturbation, de novo
ligand design, and virtual screening of available libraries. Some relevant
references are: Jorgensen, W. L.; Ruiz-Caro, J.; Tirado-Rives, J.;
Basavapathruni, A.; Anderson, K. S.; Hamilton, A. D. “Computer-aided design of
non-nucleoside inhibitors of HIV-1 reverse transcriptase.” Bioorg. Med. Chem. 2006, 16, 663-667. Tirado-Rives, J.; Jorgensen, W. L. “Contribution of
Conformer Focusing on the Uncertainty in Predicting Free Energies for
Protein-Ligand Binding.” J.
Med. Chem. 2006, 49, 5880-5884.
eCheminfo Community of Practice
Footnote:
Campus Map for direction guidance on arrival:
Download campus_mapnd.pdf
Structural Biology eCheminfo cheminformatics chemoinformatics bioinformatics Medicinal Chemistry Computational Chemistry Virtual Screening Docking Molecular Modelling Molecular Modeling pharmaceutical pharma meeting workshop training Oxford Critical Path toxicology Bursary Life Sciences Pharma Drug Discovery Research and Development Drug Development Healthcare Innovation Knowledge Management events
Comments