Second Annual Symposium
Property-Based Drug Design
Improving the Drug Discovery Process by Optimizing Bio-Physical Properties
April 15, 2013
The optimization of physical properties of a compound is fundamental to the drug discovery process, mainly due to their influence on absorption and distribution in vivo. They provide insight into the in vivo transport processes and knowing the properties will help with choosing the optimal compounds for the task. It saves costs and time when compounds are being properly analyzed in the design stage before they are moving into development, as it is important to consider questions such as how hydrophobicity will affect the solubility of a drug down the line or how the charge of the compound interacts with the absorption by a transport mechanism. Also, the use of predictive models is important, but again, without consideration of the actual physical chemical property of the new compound, the analyses will be based on a different set of data. This one-day symposium will discuss what it takes to create selective and efficacious compounds and to understand the biological data by analyzing the physicochemical properties early on.
8:30 Opening Remarks
Terry Stouch, Ph.D., President, R&D, Science for Solutions, LLC
8:40 Integrating Physicochemical Properties and PBPK for Improved Decision Making
Jan L. Wahlstrom, Ph.D., Principal Scientist, Pharmacokinetics and Drug Metabolism, Amgen, Inc.
Physicochemical properties and early ADME assays guide chemotype evaluation and rational scaffold alteration. This presentation will focus on the integration of these approaches with physiologically based pharmacokinetic modeling (PBPK) to enable the prediction of clinical outcomes and to optimize selection of development candidates.
9:20 In Silico Predictions of Ames Activities: The Nitrenium Hypothesis and Experiences with Crowd Computing
Jörg Bentzien, Ph.D., Scientist, Boehringer Ingelheim Pharmaceuticals, Inc.
This talk focuses on an ab initio approach to predict Ames activities of primary aromatic amines and recent experiences with a test example for predicting Ames activity via crowd computing using the Kaggle platform.
10:00 Morning Coffee Break
10:30 Important Considerations in the Interpretation of Pharmaceutical Data
Terry Stouch, Ph.D., President, R&D, Science for Solutions, LLC
Pharmaceutical drug discovery data can be surprisingly complex. Most of it is meant to be used immediately, in context with other temporally-related data, and with ready access to the informed commentary of the data provider. However, most companies have been archiving this data for years and it is often drawn on for use in development of predictive sciences and as as feedstock for "big data" efforts. Frequently, data is accumulated from diverse sources. However, few data are interpretable in isolation. Often involved meta data is essential to understand what a data item really means and how it relates to seemingly similar data. An especially important concern is that the precision of data is often overestimated by users. Actual error of the data items can be many times expected and may be sufficient to obviate the data for many uses. Along with examples, we will discuss issues to consider in the interpretation and use of data. We will discuss meta data of importance, magnitude and sources of error and resulting consequences for use, pitfalls in the use of historical data and accumulation of data from diverse sources.
11:00 A Thermodynamic Approach to the Optimization of Drug Candidates
Ernesto Freire, Ph.D., Henry Walters Professor, Biology and Biophysics, Johns Hopkins University
The affinity and selectivity optimization of drug candidates is difficult because it needs to simultaneously maintain or improve the drug-like properties of the compound. Recently, different metrics have been proposed to assess the quality of drug candidates. Among them the LipE or lipophilic ligand efficiency has become widely used. High quality compounds, those with a large LipE, are essentially those that derive their binding affinity from factors other than hydrophobicity. Unfortunately, LipE alone only provides a retrospective characterization of a series of compounds. It would be ideal to develop the ability to predict LipE prospectively. Thermodynamic optimization plots (TOPs) provide such a tool since LipE is proportional to the enthalpy/entropy balance of a compound. TOPs provide an easy way of organizing enthalpy, entropy and binding affinity data obtained by isothermal titration calorimetry (ITC). While traditional structure/activity relationships rely solely on binding affinity, TOPs expand the range of correlations to enthalpic and entropic coordinates. TOPs allow prediction of the enthalpic and entropic consequences of chemical modifications at specific locations in a compound. As such, it provides a way to predict the effects of specific modifications on LipE.
11:30 Identifying Good Data for Structure-Based Design
Gregory Warren, Ph.D., OpenEye Scientific Software, Inc.
Structure-based design requires protein or protein-ligand structure data. This presentation discusses how to select data with the highest quality. This is important because the quality of the data has a direct impact on the quality of the prediction.
12:00 pm In silico Predictions of Metabolism
Marvin Waldman, Ph.D., Research Fellow, Simulations Plus, Inc.
This talk discusses the development and application of Artificial Neural Network Ensembles to model inhibitors, substrates, metabolites, and kinetic constants of multiple CYP isoforms. Results will be discussed, including examples where apparently errant predictions were subsequently borne out by experiments.
12:30 Lunch on Your Own
2:00 Panel Discussion with Speakers: Considering Physicochemical Properties
• What drug properties are essential and when should they be determined?
• In silico vs experiment – is one set of data sufficient?
• Hydrophobicity in drug discovery – measurement or calculation?
• What’s next for predictive methods?
3:00 Breakout Discussion Tables
These are moderated discussions with brainstorming and interactive problem solving, allowing conference participants from diverse backgrounds to exchange ideas, experiences, and develop future collaborations around a focused topic.
Moderators:
Terry Stouch, Ph.D., President, R&D, Science for Solutions, LLC
Jan L. Wahlstrom, Ph.D., Principal Scientist, Pharmacokinetics and Drug Metabolism, Amgen, Inc.
3:50 Q&A with Speakers
4:00 End of Symposium