We are using this event page to upload some information - past and present - on the eCheminfo drug design workshop activities in Oxford, including our expansion of the workshop activity to negelected diseases.
As in previous workshops the emphasis is on problem solving, practical hands-on use of methods and software applied to drug discovery problems, and working together throughout the week. The workshop will also offer a neglected disease case study focus.
I heard the sad news of your sudden death today Patrik. We will miss your science, your fun and your company. We will remember the good times. Below we are setting out on what became a tradition at eCheminfo summer workshops in Oxford: the start of the week evening punting trip, this one in 2009.
I will deliver an OpenTox seminar at the IQPC (Pharma IQ) Predictive Toxicology conference taking place in London 23-24 Feb 2011, in which I will present on recent developments with regards to OpenTox interoperability advances, and particularly to support of collaborative drug discovery projects. Also of interest to the increasing support by OpenTox of in vitro methods is a significant program on stem cell technology. The conference will address the accuracy of drug safety testing through the integration of advanced prediction methods, including stem cell based assays and screening into safety evaluation strategies. The main conference will take place in London on 23rd-24th February 2011 and new for this year, there will be a pre-conference focus day on 22nd February 2011 dedicated to the use of stem cells as a tool to enhance predictive toxicology. http://www.predictivetoxicology.co.uk/Event.aspx?id=394530
During the workshop we will apply a variety of design and modelling methods to drug discovery problems guided by workshop leaders with expertise in the approaches used. A case study approach will additionally be followed so that groups can work together throughout the week on their case study problems. The case studies will also be developed virtually before the workshop week with support extended afterwards for further work including experimental testing of interesting results and hypotheses developed. Case studies will include neglected disease targets and challenging kinase design problems.
A combination of design techniques including pharmacophores, ligand-based screeening and docking will be applied to problems. Additionally the use of consensus methods and prediction rules to optimally combine predictions from multiple methods will be studied. Prediction results across structures and methods will be displayed across a collaborative dashboard for interactive library design. ADME and toxicology predictions will be carried out for library structures and incorporated in decision rules. The use of quantum mechanical methods to more accurately model ligand and protein interactions and refine scoring will be studied. Fragment-based approaches will be applied to the design of molecules with promising chemistries. Physiologically-based simulations will be used to study the pharmacokinetic properties of structures and their application to lead optimisiation.
Through the time spent working and discussing together combined with the availability of a variety of leading software and expert support from workshop leaders, workshop participants should take home ideas and learning to help accelerate their own projects. The location and atmosphere in Oxford is also an ideal background for networking, getting to know your peers and joining the ongoing eCheminfo community of practice activities. As is often common with eCheminfo gatherings the workshop usually attracts a variety of backgrounds including industry, academia and government research instititutes and from many different countries. Non-cheminformatics specialists from different areas of chemistry and biology often participate and bring an interdisciplinary interaction to the groups.
Monday July 26 08.30 Registration Open 09.00 Overview of Workshop Activities, Presented by Barry Hardy (Douglas Connect) We will review the approach to the workshop including group work and case studies. 09.15 Insights into Kinase Structures and Ligand Design, led by Jeffrey Wiseman (Pharmatrope) and Barry Hardy (Douglas Connect) An overview of design strategies for kinase targets will be presented and discussed including:
- Analysing the Kinome and Kinase Families - Biology of Human and Parasitic Kinases - Target Selection & Validation - Modelling Target Structures - Kinase Inhibition Features and Ligand Design Strategies - Addressing Potency and Selectivity Issues
12.00 Lunch 13.00 Structure-focused Pharmacophores for the Identification of Novel, Led By Gerhard Wolber (Inte:Ligand) Virtual screening using 3D pharmacophores has been established as an important and commonly used technique for virtual screening in recent years . Using several current biological targets we demonstrate methods supported by the software tool (LigandScout), which allows for rapid and transparent development of high-quality 3D pharmacophores. Besides the easy and automated creation of such models from protein/ligand complexes, LigandScout provides intuitive pharmacophore overlay and interpolation workflows based on a robust and fast chemical-feature-based alignment algorithm. The underlying methods are scientifically published [2-4] and based on several years of experience in pharmacophore creation. The full-featured 3D graphical user interface makes the exploration of the active site and pharmacophore creation within the protein-ligand complex user-friendly and transparent. Binding site analysis, pharmacophore-based alignment and the creation of shared feature models are designed to make the drug discovery workflow more efficient. In this half-day workshop we will demonstrate how to derive structure- and ligand-based pharmacophores using the LigandScout modelling environment and how to use them for fast and accurate virtual screening. Model validation  and predictivity assessment will include Receiver Operating Characteristic (ROC) curves and enrichment analyses. Class members working in small groups will be able to apply the pharmacophore methods during the week to their case studies.
References  Langer T., Hoffmann R. D. Pharmacophores and Pharmacophore Searches; VCH/Wiley, Methods and Principles in Medicinal Chemistry, Vol. 32, R. Mannhold, H. Kubinyi, G. Folkers, series editors, ISBN: 3527312501 (2006)  Wolber G., Seidel T., Bendix, F., Langer T. Molecule-pharmacophore superpositioning and pattern matching in computational drug design;.Drug Discovery Today. 2008 Jan ;13 (1 2):23-9.  Wolber, G.; Dornhofer, A. A.; Langer, T.; Efficient overlay of small organic molecules using 3D pharmacophores; J. Comput. Aided Mol. Des.; 2007; 20(12); 773-788.  Wolber, G.; Langer, T.; LigandScout: 3-D Pharmacophores Derived from Protein-Bound Ligands and Their Use as Virtual Screening Filters; J. Chem. Inf. Model; 2005; 45(1); 160-169.  Kirchmair, J. et al. Enhancing drug discovery through in silico screening: strategies to increase true positives retrieval rates. Curr. Med. Chem.; 2008; 15; 2040-2053.
16.00 Group Work and Discussion on Workshop Case Study Problems 18.00 Poster Session with Refreshments and Food
Tuesday July 27 08.45 Shape and Electrostatics in Virtual Screening and Lead Hopping, Led By Paul Hawkins (OpenEye) This session will focus on two tools that utilise very similar approaches for two different purposes: ROCS for shape-based virtual screening and lead-hopping and BROOD for lead-hopping and bioisostere identification using shape and electrostatics. In the ROCS section the workshop participant will learn about searching in shape space, using shape in virtual screening and the visual query editor vROCS. The vROCS editor will be used to generate queries for virtual screening or lead-hopping experiments. Using vROCS we will learn how to merge multiple molecules into a single query, edit molecules in a way that separates the structure of a molecule from the idea of a query, and validate the queries that class members generate in retrospective virtual screens. The workshop participants will learn about robust statistical methods that can be applied to virtual screening experiments and will use these methods to compare queries that they generate, enabling them to choose the best query for prospective experiments. In the BROOD workshop section, class members will learn how shape and electrostatic similarity can be applied at the fragment level (for lead-hopping and isostere replacement) and will learn how to use the BROOD Graphical User Interface (GUI). The GUI will be used to generate an isostere replacement query and to run that query against a large pre-generated database of fragments. Class members working in small groups will be able to apply the screening methods during the week to their case studies.
11.30 Group Work on Workshop Case Study Problems 13.00 Lunch
14.00 Consensus Strategies for Challenging Dockings, Led By Alessandro Contini (University of Milan) The attention toward protein-protein interactions as potential targets for the design of very specific drugs dramatically increased during the last few years. The binding regions involved in protein-protein interactions are rarely deep and well-defined cavities, as found in the active site of an enzyme, but are often characterized by solvent-exposed clefts on the protein surface. This feature makes the definition of a reliable docking protocol quite challenging.
This workshop will be focused upon the development of a docking protocol for the identification of small organic molecules potentially able to interfere with the interaction of two target proteins (hereafter referred as proteins A and B). The study will cover two possible scenarios:
1) The crystal structure of the AB complex is available; 2) The structure of either A or B is available co-crystallized with a known inhibitor.
In both scenarios we will analyze the complex, finding and fixing all possible sources of error and preparing the model for docking experiments (fixing missing residues, assigning the correct protonation state, relaxing the structure through short molecular dynamic runs followed by geometry minimizations). A docking protocol will then be realized and tested through the docking of small databases of compounds with known activity. The results of both scenarios will be compared and discussed.
As consensus between different computational methods might be a key to improve the success rate of a virtual screening procedure, an alternative docking protocol will be also realized by using a different method and software. The two protocols will be tested within a virtual screening case study and those hits identified by each single protocol will be compared to those identified by both protocols applied on a consensus basis.
16.30 Group Work on Workshop Case Study Problems 18.00 End of Workday 18.00 Punting Trip (Weather Permitting)
Wednesday July 28 8:45 Theory to Application: How Quantum Mechanics can be Applied to Structure-Based Drug Discovery, Led by Lance Westerhoff (QuantumBio) Traditionally, linear scaling, quantum mechanics-based (QM-based) methods for characterization of target/ligand complexes have been better suited to academic environments as they are sometimes difficult applications to access in the industrial domain. Recently, QuantumBio has bridged that gap through the development of several QM-based interaction profiling tools specifically tailored to the structure-based drug discovery process. When plugged into MOE, these tools – including scoring, pair-wise interaction energy decomposition, and QSAR – become better integrated with the workflows commonly used in the field. To date, this work has lead to the development of three major MOE svl plugins: MOE/QMScore, MOE/NMRScore, and MOE/QM-PWD. We are now able to prepare any number of QM simulations using the MOE graphical user interface (GUI), execute the simulations in parallel using MOE's message passing infrastructure, and finally import the results back into the MOE GUI for further analysis.
As a use case, these QM simulations have been carried out for a series of protein kinase B inhibitors derived from fragment (FBDD) and structure-based drug design (SBDD). These protein-ligand complexes were selected because they represent a consistent set of experimental data that includes both crystal structures and affinities. Seven scoring functions were constructed based on a mixture of several quantum- and molecular- mechanical methods. The optimal models obtained by statistical analysis of the aligned poses are predictive as measured by a number of standard statistical criteria and validation with an external data set. Together, this model provides residue-based contributions to the overall binding affinity, and these results are treated using both native MOE analytical methodologies and customized widgets including the QM-PWD Interaction Energy (IE) Map, Structure/Activity Relationship (SAR) Map, and results tables. The IE map highlights the most important residues for ligand binding, while the SAR Map highlights residues that are most critical to discriminating between more and less potent ligands. Taken together the Interaction Energy and SAR Maps provide useful insights into drug design that would be difficult to garner in any other way.
Class members working in small groups will be able to study together the setup of the QM methods during the workshop applicable to their case studies. As calculations can be computationally-intensive, we will offer the possibilities of running computations both before and after the workshop, made available through a collaboration environment.
11.30 Group Work on Workshop Case Study Problems 13.00 Lunch
14.00 Virtual High-Throughput Screening: A Powerful Strategy for Drug Discovery, Presented by Katie Simmons (University of Leeds and SimBioSys Inc.) In silico molecular docking techniques, such as virtual high-throughput screening (VHTS), are powerful approaches to the discovery of new enzyme inhibitors. Additionally, de-novo design is a complementary strategy for inhibitor discovery. Here, by using the structural features present within the enzyme only, new inhibitor designs are built-up sequentially according to the requirements of the targeted binding site. Therefore, de-novo design is an important technique to use in parallel with VHTS in a particular hit identification campaign, as a good de-novo design program will examine structural space larger by many orders of magnitude than that of most virtual libraries currently used for this purpose.
We have recently applied both the de-novo molecular design computer program SPROUT, and the VHTS program eHiTS to a number of therapeutically attractive enzyme targets and have, in the majority of cases under study, rapidly identified inhibitors in the micromolar range or better.
References (1) Bioorganic & Medicinal Chemistry Letters, Volume 19, Issue 23, 1 December 2009, Pages 6770-6774 (2) Discovery of novel non-peptide inhibitors of BACE-1 using virtual high-throughput screening, N. Yi Mok, James Chadwick, Katherine A.B. Kellett, Nigel M. Hooper, A. Peter Johnson, Colin W.G. Fishwick; doi:10.1016/j.bmcl.2009.09.103, J. Med. Chem. 2009, 52, 2683 - 2693 (3) Structure-Based Design, Synthesis, and Characterization of Inhibitors of Human and Plasmodium falciparum Dihydroorotate Dehydrogenases, Matthew Davies,Timo Heikkila, Glenn A. McConkey, Colin W. G. Fishwick, Mark R. Parsons, and A. Peter Johnson; Mol Microbiol. 2009 Apr;72(2):335-43. Epub 2009 Mar 3. (4) The nature of Staphylococcus aureus MurA and MurZ and approaches for detection of peptidoglycan biosynthesis inhibitors; Blake KL, O'Neill AJ, Mengin-Lecreulx D, Henderson PJ, Bostock JM, Dunsmore CJ, Simmons KJ, Fishwick CW, Leeds JA, and Chopra I.
14:30 User-friendly Ligand-based Filtering and Docking: Results Analysis and Visualization, Led by Katie Simmons (University of Leeds and SimBioSys Inc.) Participants in this workshop will experience applying different methods and tools that can make virtual screening more productive. eHiTS LASSO is a ligand-based filter that uses the chemical features of a ligand surface to create a pseudo-pharmacophore for rapid screening of large databases. eHiTS is an accurate fragment-based docking program for both virtual screening and binding pose prediction of ligands. CheVi is an advanced visualization package specifically designed to help users analyze how ligands interact with receptors. In addition, CheVi acts as a front-end graphical user interface to run eHiTS LASSO and eHiTS screening and docking jobs.
The workshop will describe a typical work-flow from database to lead and show how tools from SimBioSys can make the process more effective using the plasmodium falciparum dihydroorotate dehydrogenase inhibitor, A77-1726 as an example. Users will get a hands-on lesson on how to use all the tools described above, with specific attention to analysis of interaction results.
Class members working in small groups will be able to apply the virtual screening methods during the week to their case studies.
Participants will also take home: A free unlimited version of the CheVi visualization tool and a two month evaluation version of eHiTS and the eHiTS LASSO.
16.45 Group Work on Workshop Case Study Problems 18.00 End of Workday
Thursday July 29 8.45 Fragment-Based Ligand Design: Teaming up Medicinal and Computational Chemists, Led by Peter Oledzki (BioSolveIT) Lead discovery often starts from small fragment binders for which experimental evidence has been found in an active site. Development into a lead structure can involve three possible scenarios: a) to grow from these 'needles' into the depths of the pocket; b) merging multiple overlapping binders into a single potent lead; or c) the more difficult prospect of linking two or more fragments into one compound with optimized potency.
These tasks can now be accomplished computationally with a novel software tool, LeadIT, which was primarily designed for mixed medicinal and computational chemistry teams. Synthetically accessible compounds can be generated in seconds using fragment based design by using an indexed 3D fragment library of fragments. We will elucidate the basic principles of the approach and give examples which map onto experimental data and evolve into novel lead ideas. Workshop participants may then proceed to working on individual hands-on exercises and application of the methods to their case study problems.
11.30 Group Work on Workshop Case Study Problems 13.00 Lunch
14.00 Driving Lead Optimisation through exploitation of pharmacokinetic properties predicted from early ADME inputs: a role for physiologically-based pharmacokinetic (PBPK) modelling , Led by Simon Thomas (Cyprotex) Physiologically based pharmacokinetic (PBPK) modelling provides a powerful means of integrating ADME and physicochemical data to predict in vivo pharmacokinetics in humans and pre-clinical animals. Predictions of pharmacokinetics (PK) from ADME data can enhance the ability to select compounds that are most likely to have appropriate PK in vivo. The determination of physicochemical and ADME properties during early drug discovery ('early ADME data') enables PK prediction to be performed at any stage from lead identification onwards. PK prediction thus serves to integrate the data from various ADME/physchem screens – whether in vitro or in silico – greatly increasing their value over and above that of the raw data alone. In particular the role of sensitivity analysis – in which the effect of uncertainty in an input property on the value of an output (predicted) property is quantified – is a powerful tool for informing, and helping to direct – chemistry during lead optimisation.
In this workshop, the focus will be on understanding the fundamentals of PBPK modelling, the use of appropriate ADME and physicochemical data as inputs, and the utilisation of results during early drug discovery. For case study investigation of various aspects of PK prediction, participants will have access to Cloe® PK and Cloe Predict® HIA (Human Intestinal Absorption) software. These are powerful, yet intuitive, web-based programs using PBPK models for PK prediction. Their simple data inputs and comprehensive reporting make them suitable, not only for ADME/PK scientists, but also for medicinal chemists and biologists. In addition to the prediction of PK properties, the use of univariate and multivariate sensitivity analyses as an aid to directing chemistry optimisation will be investigated.
16.30 Group Work on Workshop Case Study Problems 18.00 End of Workday
Friday July 30 9.00 Experimental Testing of Predictions We will review the experimental design and planning for the testing of predictions. 10.00 Group Work on Workshop Case Study Problems 12.00 Group Presentation of Workshop Case Study Results 13.00 Lunch
14.00 Group Work on Workshop Case Study Problems 16.00 Group Presentation of Workshop Case Study Results
During April - October 2010 we will be running two pilots which will create two virtual organisations (VOs) of research partners who will collaborate on a drug discovery and a predictive toxicology project.
We are currently seeking Expressions of Interest in the ICT support role for the pilots which should be submitted by 17.00 CET, 19 February 2010 through the following form: https://www.surveymonkey.com/s/QSHKDCH
The VOs will be created from partners from Douglas Connect Collaboration Pools, who will bring contributing knowledge, computational or experimental capabilities to the partnership for the duration of the projects:
a) Drug Design project will examine the application of a variety of leading modelling and design approaches to novel target kinases, to experimentally test predictions, and to initiate a best practices virtual screening resource.
b) Predictive Toxicology project will apply a combination of in silico and in vitro approaches to predict in vivo toxicity including exposure.
The ICT support role will support the informatics requirements of the researchers in organising, processing and sharing data, and in collaborative analysis and decision making.
The VOs will be supported by knowledge-oriented collaboration services developed under the FP7 Synergy research project, including a reactive complex event driven engine, collaboration moderator, collaboration pattern services and partner knowledge base. Summary on Synergy project: http://barryhardy.blogs.com/theferryman/2008/02/knowledge-orien.html
The Predictive Toxicology VO will be supported by distributed REST-driven web services for data management, model building, validation and reporting, developed under the OpenTox Framework (http://www.opentox.org/)
Barry Hardy PhD Director, Community of Practice & Research Activities and OpenTox Project Coordinator (www.opentox.org) Douglas Connect Switzerland
We are establishing an eCheminfo Collaboration Pool of individuals and organisations who have an interest in collaborating together in areas of drug discovery. For example, a selection of Pool members could participate in a collaborative drug discovery or predictive ADME/Tox project, develop a funding proposal or response to a call opportunity together, or develop an innovation or best practice.
Collaborations will take a virtual organisation approach, i.e., partners can bring contributing modelling or experimental capabilities to a partnership for the duration of a project or other endeavour.
We will run two such virtual organisation (VO) projects for the first time this year starting later this Spring: A) Drug Design project which will examine the application of a variety of leading modelling and design approaches to novel target kinases, to experimentally test predictions, and to initiate a best practices virtual screening resource. B) Predictive Toxicology project to apply a combination of in silico and in vitro approaches to predict in vivo toxicity including exposure.
Both VOs will be supported by a variety of leading modelling and design software, informatics infrastructure, and collaborative content management and electronic lab notebook systems.
The VOs will be supported by knowledge-oriented collaboration services developed under the FP7 Synergy research project, including a reactive complex event driven engine, collaboration moderator, collaboration pattern services and partner knowledge base.
The Predictive Toxicology VO will be supported by distributed REST-driven web services for data management, model building, validation and reporting, developed under the OpenTox Framework (http://www.opentox.org/).
We will also consider incorporation of Case Studies into the eCheminfo Drug Discovery and Predictive ADME/Tox workshops to be held in Oxford this summer. See http://echeminfo.com/
Barry Hardy PhD Director, Community of Practice & Research Activities Douglas Connect Switzerland