THURSDAY, JUNE 9
8:00 am Morning Coffee
8:25 Chairperson’s Opening Remarks
Woody Sherman, Ph.D., Vice President, Applications Science, Schrodinger, Inc.
8:30 FAST Prediction of Protein Stability and Flexibility
Donald Jacobs, Ph.D., Associate Professor of Physics, Physics and Optical Science, University of North Carolina
Funded by NIH R01-GM073082, a novel computational method based on free energy decomposition and an iterative self-consistent reconstitution involving network rigidity to account for non-additivity in conformational entropy due to correlated motions has been developed. The program, FAST, provides a Flexibility And Stability Test for proteins that quantifies stability/flexibility relationships for understanding function. Speed versus accuracy is optimized for high-throughput screening to test for structure/function relationships in mutation studies, substrate binding, pKa predictions, allostery, protein-protein interactions and protein formulation.
8:55 Predicting Selectivity and Druggability in Small Molecule Drug Discovery
Alan Cheng, Ph.D., Senior Scientist, Chemistry Research & Development, Amgen, Inc.
Structure-based druggability and selectivity analyses increasingly inform target assessment and setting of lead optimization strategies in drug discovery. Continuing work on druggability prediction will be presented in the context of applying druggability prediction approaches to large numbers of protein structures. Work on selectivity analysis will be presented in the context of kinase inhibitor discovery examples from Amgen, and we will show how a thermodynamics-based selectivity index can help in predicting biological selectivity.
9:20 Development of a Simple Metric for the Structure-Based Assessment of Protein Druggability
Emanuele Perola, Ph.D., Research Fellow I, Vertex Pharmaceuticals
Accurate assessments of target druggability could significantly reduce the attrition rate in drug discovery programs. We developed an algorithm to isolate and characterize the binding pockets of protein targets and used it to analyze a set of validated drug targets and a diverse set of proteins with known crystal structures. We identified five parameters that differentiate the two sets in a statistically significant manner and derived a set of simple rules that could be applied to estimate the druggability of prospective targets.
9:45 Improvements in Predicting Protein-Ligand Binding Energies: Waters, Protein Flexibility, and Empirical Scoring
Woody Sherman, Ph.D., Vice President, Applications Science, Schrödinger, Inc.
Recent years have seen significant advancements in structure-based drug design methods. The inclusion of explicit waters and protein flexibility have proven to be critical components of improved protein-ligand scoring. We will discuss these advances and how they have produced more robust scoring and have been integrated into the Glide XP empirical scoring function to significantly improve virtual screening enrichments.
10:15 Networking Coffee Break in the Exhibit Hall with Poster Viewing
10:45 Identification of Druggable Sites for Protein-Protein Interaction Targets by Computational Fragment Mapping
Dima Kozakov, Ph.D., Research Assistant Professor, Department of Biomedical Engineering, Boston University
We have developed computational fragment mapping to identify “hot spot” regions in protein-protein interfaces. The method accounts for protein plasticity, and finds energetically favorable sites for fragment sized probe molecules. Results are presented for protein-protein interaction targets, including interleukin-2, Bcl-xL, MDM2, HPV-11 E2, ZipA, TNF-a, NEMO, and eIF4E. We also discuss methods of finding druggable targets in cancer pathways.
11:10 Induced Fit Modeling with Monte Carlo Techniques: Protein Energy Landscape Exploration
Victor Guallar, Ph.D., ICREA Research Professor, Life Science Department, Barcelona Supercomputing
Protein energy landscape exploration (PELE) constitutes a remarkable advance over conventional techniques to map protein and protein-ligand dynamics. This method, which combines protein structure prediction techniques with a metropolis algorithm, is capable of describing the all-atom ligand migration pathway and induced fit in approximately 100 hours of CPU. A more recent developments in protein-ligand and protein-protein induced fit will be presented.
11:35 Antagonizing the IAP proteins to Induce Programmed Cell Death in Cancer: A New Therapeutic Approach Utilizing Protein-Protein Interaction
Kurt Deshayes, Ph.D., Senior Scientist, Department of Early Discovery Biochemistry, Genentech, Inc.
Inhibitor of apoptosis (IAP) proteins are expressed at elevated levels in human malignancies and block cell death in response to diverse stimuli. They are targets for small molecule cancer therapeutics under current clinical development. Results from biophysical, chemical, and molecular biological studies of the mechanism by which binding of small molecule IAP antagonists leads to cancer cell death will be reported.
Sponsored by12:00 pm Target based lead optimization: VLifeSCOPE and SATREASudhir A. Kulkarni, Ph.D., Vice President, Discovery Research, VLife Sciences Technologies Pvt. Ltd.VLifeSCOPE is a novel target based lead optimization method where target ligand interactions obtained from scoring function are partitioned into individual residues in the active site and correlated to the activity of molecules. VLifeSCOPE provides information on key residues to be targeted in lead optimization.
SATREA is a visualization tool that provides information on regions of exploitation and avoidance for achieving specificity of target of interest with respect to other target in the same family.
12:30 Luncheon Presentations (Sponsorship Opportunities Available) or Lunch on Your Own
1:30 Chairperson’s Remarks
Roderick Hubbard, Ph.D., Professor, University of York and Vernalis Ltd.
1:35 Practical Concepts in Fragment-to-Lead Optimization
Marcel Verdonk, Ph.D., Director, Computational Chemistry and Informatics, Astex Therapeutics Ltd.
Fragment-based screening methodologies have become widely used in drug discovery projects. We will present various practical concepts that are used at Astex during the process of fragment prioritization and optimizations. For example, we will discuss the use of Ligand Efficiency and Group Efficiency measures. In addition, a practical approach will be presented to significantly improve docking performance during fragment-to-lead optimization
2:00 Making Decisions in Fragment-Based Discovery
Roderick Hubbard, Ph.D., Professor, University of York and Director, Structural Sciences, Vernalis Ltd.
Fragment-based methods are now well established and generate hits for most targets. A continuing challenge is making the right decisions during the fragment to lead evolution stage – which fragment to evolve based on chemical and target opportunities and how to evolve fragments in the absence of structure. This presentation will discuss the contribution of both computational and experimental methods to address these challenges.
2:30 Applying Integral Equation Theory to Structure Based Design
Jean-François Truchon, Ph.D., Research Scientist, Chemical Computing Group
Integral equation theories seem attractive for their ability to generate water and solvated fragment density distributions in enzyme active sites. They are much less computationally demanding than Molecular Dynamics and unveil explicit solvent details missing in continuum models such as Poisson-Boltzmann. We compare a particular flavor, namely 3D-RISM/KH, to other methods and examine its usefulness in structure based design and fragment docking.
3:00 Networking Refreshment Break in the Exhibit Hall with Poster Viewing
3:45 A Computational Approach to Fragment-Based Drug Design
Charles H. Reynolds, Ph.D., Senior Director, Discovery Technologies, Ansaris
We have developed a free energy method for computing the interaction of small molecule fragments with their target proteins. This approach is fast relative to traditional free energy methods and more accurate than traditional modeling methods that only compute interaction energies based on one, or a small number, of configurations. The methodology is ideal for characterizing critical waters (e.g. tight-binding), assessing the affinities of small molecule fragments, and identifying hot spots in protein-protein interaction targets.
4:00 Next-Generation Shape Signatures: A Powerful Tool for Fragment-Based Drug Discovery
Randy Zauhar, Ph.D., Associate Professor, Chemistry & Biochemistry, University of the Sciences
The original Shape Signatures method uses a ray-tracing approach to efficiently explore the volume and surface properties of a molecule. In our new approach, molecules are automatically partitioned into fragments, and the Shape Signatures descriptors are now likewise fragment-based. Query and target molecules are now compared by matching fragment in all ways compatible with the underlying structure.
4:25 Panel Discussion
Topic: What are the challenges and opportunities for the next stages of development in fragment-based methods?
For many targets, the identification of fragments that bind is for the most part a solved problem. There are two main issues for the methods. The first is deciding which fragments to take forward – there can be hit rates of 5-10% for some targets. The second is how to guide the early stages of chemical optimisation in the absence of crystal structures, which is a particular problem for challenging targets such as protein-protein interactions. The discussion will ask for opinions on how computational methods could tackle these issues and what developments are still needed.
5:00 Networking Cocktail Reception in the Exhibit Hall with Poster Viewing
6:00 End of Day