2020 Workshop on Free Energy Methods in Drug Design

Virtual Conference 15-17th June 2021

Registration is open! Register Here

Poster session hosted by Schrodinger

PDT	ET	UK	CET	Tuesday 15th June 2021
8.15am	11.15am	4.15pm	5.15pm	Introduction from organising committee
				Session I - moderator: Antonia Mey
8.30am	11.30am	4.30pm	5.30pm	Julien Michel (Edinburgh) - Adaptive absolute binding free energy calculations for lead compounds and flexible proteins recorded
8.55am	11.55am	4.55pm	5.55pm	Hannah Baumann (UC Irvine, Mobley lab) - Binding free energy calculations: Sampling challenges, nonequilibrium switching, and alternate approaches recorded
9.20am	12.20pm	5.20pm	6.20pm	Emilio Gallicchio (Brooklyn CUNY) - The Alchemical Transfer Method: a Streamlined Approach to Absolute and Relative Binding Free Energy Calculations recorded
9.45am	12.45pm	5.45pm	6.45pm	break
				Session II -moderator: Sereina Riniker
10.00am	1.00pm	6.00pm	7.00pm	Christopher Bayly (Openeye) - Non-Equilibrium Switching in Orion
10.25am	1.25pm	6.25pm	7.25pm	Michael Shirts (Boulder) - Learning Free Energy Surfaces recorded
10.50am	1.50pm	6.50pm	7.50pm	POSTERS
				Session III - moderator: Jonah Vilseck
11.15am	2.15pm	7.15pm	8.15pm	David Minh (Illinois) - Large-scale Free Energy Calculations with Implicit Ligand Theory recorded
11.40am	2.40pm	7.40pm	8.40pm	Cindy Yan (Merck) - Binding affinity predictions in small-molecule drug discovery: Is balancing accuracy and efficiency possible? recorded
12.05pm	3.05pm	8.05pm	9.05pm	Matteo Aldeghi (Vector Institute) - Structure-based predictions of kinase inhibitor resistance: a prospective evaluation recorded

PDT	ET	UK	CET	Wednesday 16th June 2021
8.15am	11.15am	4.15pm	5.15pm	Introduction
				Session I-moderator: Richard Lonsdale
8.30am	11.30am	4.30pm	5.30pm	Benjamin Ries (ETH Zurich, Riniker lab) - Relative binding free energies with scaffold-hopping type transformations using RE-EDS recorded
8.55am	11.55am	4.55pm	5.55pm	Jonathan Essex (Southampton) - Combining grand canonical and nonequilibrium candidate Monte Carlo to enhance sampling
9.20am	12.20pm	5.20pm	6.20pm	Wonpil Im (Lehigh) - What can CHARMM-GUI do for you? recorded
9.45am	12.45pm	5.45pm	6.45pm	break
				Session II -moderator: Zoe Cournia
10.00am	1.00pm	6.00pm	7.00pm	Jonah Vilseck (Indiana) - Multisite Sampling with Discrete Gibbs Sampler-Based Lambda-Dynamics recorded
10.25am	1.25pm	6.25pm	7.25pm	William Jorgensen (Yale) - 40 Years of Free-Energy Calculations from Solvent Effects to SARS-CoV-2 Inhibitors recorded
10.50am	1.50pm	6.50pm	7.50pm	POSTERS
				Session III -moderator: Emilia Pecora de Barros
11.40am	2.40pm	7.40pm	8.40pm	Yutong Zhao (Relay Tx) - Improved Orientational Restraints for Binding Free Energies
12.05pm	3.05pm	8.05pm	9.05pm	Lyna Luo (Western) - Understand and optimize reversible covalent inhibitors and allosteric agonists recorded

PDT	ET	UK	CET	Thursday 17th June 2021
8.15am	11.15am	4.15pm	5.15pm	Introduction
				Session I-moderator: Camilo Velez-Vega
8.30am	11.30am	4.30pm	5.30pm	Niels Hansen (Stuttgart) - Exploring a correlation between Soret and partition coefficient by means of free-energy calculations recorded
8.55am	11.55am	4.55pm	5.55pm	Robert Abel (Schrodinger) - Models to Medicines: The impact of large scale physics-based design on clinical candidate discovery recorded
9.20am	12.20pm	5.20pm	6.20pm	Vytautas Gapsys (Max Planck Institute) - Alchemical Absolute Protein-Ligand Binding Free Energies for Drug Design recorded
9.45am	12.45pm	5.45pm	6.45pm	break
				Session II -moderator: Kira Armacost
10.00am	1.00pm	6.00pm	7.00pm	Joy Yang (Pfizer) - Optimal Designs for Pairwise Calculation: An Application to FEP in Minimizing Prediction Variability
10.25am	1.25pm	6.25pm	7.25pm	Brian Radak (Silicon Tx) - Advancing Drug Discovery with a Practical Approach to Free Energy Simulations
10.50am	1.50pm	6.50pm	7.50pm	POSTERS
				Session III -moderator: Zhuyan Guo
11.15am	2.15pm	7.15pm	8.15pm	Dave Thompson (CCG) + Kira Armacost(GSK) - "Free Energy Methods in Drug Discovery : “It will be the most important field in computational chemistry.”" - recorded
11.40am	2.40pm	7.40pm	8.40pm	Gary Tresadern (Janssen)
12.05pm	3.05pm	8.05pm	9.05pm	Gary Tresadern (Janssen)

POSTERS Tuesday

Some presentations can be viewed ahead of the meeting here

PDT	ET	UK	CET	Tuesday 15th June 2021	Abstract
10.50am	1.50pm	6.50pm	7.50pm	POSTERS
				Michael Gillhofer BOKU University of Natural Resources and Life Sciences, Vienna	Importance of water molecules for alchemical free-energy calculations: One challenge in the field of free-energy calculations is the occurrence of water molecules since inappropriate placements can lead to a stabilization of different conformational orientations of the ligand. With classical alchemical perturbation methods such as thermodynamic integration (TI), it is essential to know the amount of water molecules in the active site of the respective ligands, but the resolution of the crystal structure and the correct assignment of the electron density can sometimes lead to challenges in the placement of water molecules. This work aims to calculate relative binding free-energy differences using extended TI for two different ligands for the system of the heatshock protein 90 (HSP90), where it has been shown that the correct amount of water molecules in the active site can change with the respective ligand. To determine the correct number of waters a newly developed technique AEDS (accelerated enveloping distribution sampling) was used. This technique allows the perturbed waters to fluctuate between dummy and real state, providing information about the number of waters preferred for the ligand of interest.
				Oriol Gracia i Carmona Universität für Bodenkultur Wien (BOKU)	One of the most challenging aspects in the calculation of free energies is the presence of slow relaxing processes. The reason behind this is that the state-of-the-art methods currently used rely on performing several short simulations, increasing the number of times that those slow processes need to be sampled. Such sampling limitation is often overcome by adding a bias based on previous knowledge of the system, which may not be always available. AEDS (Accelerated enveloping distribution sampling) is a recently developed technique that allows efficient sampling of the end-states in a single simulation which integrates the use of a reference state with accelerated MD. With the aim of testing the capabilities of AEDS and of providing guidelines on how to setup and analyze AEDS simulations, we have used T4 L99A lysozyme as a test system. The T4 L99A lysozyme contains a valine, Val 111, which conformation highly influences the binding efficiencies of the ligands, making it an ideal test system to see the performance of AEDS on a slow relaxing process.
				Jonathan Barnes University of Idaho	Analysis of Software Methods for Estimation of Protein-Protein Relative Binding Affinity: A growing number of computational tools have been developed to accurately and rapidly predict the impact of amino acid mutations on protein-protein relative binding affinities. Such tools have many applications, for example, designing new drugs and studying evolutionary mechanisms. In the search for accuracy, many of these methods employ expensive yet rigorous molecular dynamics simulations. By contrast, non-rigorous methods use less exhaustive methods, allowing for more efficient calculations. This trade-off between accuracy and computational expense makes it difficult to determine the best method for a given context. Here, eight non-rigorous computational methods were assessed using eight antibody-antigen and eight non-antibody-antigen complexes for their ability to accurately predict relative binding affinities (ΔΔG) for 654 single mutations. Our results suggest these methods can be used to quickly and accurately predict stabilizing versus destabilizing mutations but are less accurate at predicting actual binding affinities.
				Anthony Hazel University of Maryland, Baltimore	Rapid and accurate estimation of protein–ligand relative binding affinities using site-identification by ligand competitive saturation: Predicting relative protein–ligand binding affinities is a central pillar of lead optimization efforts in structure-based drug design. The site identification by ligand competitive saturation (SILCS) methodology is based on functional group affinity patterns in the form of free energy maps that may be used to compute protein–ligand binding poses and affinities. Using the SILCS methodology we can achieve an average rank-ordering accuracy of 74-77% when considering eight well-studied protein. This accuracy increases to 80-82% when the SILCS atomic free energy contributions are optimized using a Bayesian Markov-chain Monte Carlo approach. Overall, the SILCS methodology yields similar or better-quality predictions without a priori need for known ligand orientations in terms of the different metrics when compared to current alchemical-based approaches with significant computational savings while additionally offering quantitative estimates of individual atomic contributions to binding free energies. These results further validate the SILCS methodology as an accurate, computationally efficient tool to support lead optimization and drug discovery.
				Sheenam Khuttan Brooklyn College, CUNY	Relative Binding Free Energy Calculations for the GDCC SAMPL8 Blinded Challenge using the Alchemical Transfer Method: Alchemical Relative Binding Free Energy (RBFE) calculations are increasingly becoming one of the work horses of structure-based drug development. However, applications involving variations of ligand charge and ligand scaffold, such as ring opening/closing and linker deletion/extension, remain challenging. Here, we illustrate the application of our Alchemical Transfer Method (ATM) to the calculation of relative binding free-energy differences for the Gibbs Deep Cavitand Complexes (GDCC) SAMPL8 host-guest challenge benchmark. ATM is applied to the said complexes, each with one ligand at the binding site and the second unbound ligand in water, all in a single solvent box. The application of an alchemical perturbation switches the initial positions of the two ligands through a symmetric alchemical intermediate. We found that ATM's dual topology protocol can successfully tackle ligand pairs with different scaffolds and ionization states. We continue to illustrate applications of ATM RBFE to protein-ligand systems with different scaffolds.

POSTERS Wednesday

Some presentations can be viewed ahead of the meeting here

PDT	ET	UK	CET	Wednesday 16th June 2021	Abstract
10.50am	1.50pm	6.50pm	7.50pm	POSTERS
				Jenke Scheen University of Edinburgh	FEP-Space: Towards Data-driven estimation of optimal perturbation networks in relative alchemical free energy calculations: Alchemical free energy (AFE) calculations are helpful in solving the ligand-optimisation problem in both academic and corporate drug discovery. A critical step in relative AFE workflows is the generation of effective perturbation networks to ensure transformations are being simulated with sufficient phase-space overlap. Currently, state-of-the-art softwares use primarily molecular similarity metrics to estimate optimal edges in networks. This method often still requires the user to review and tweak the presented perturbation network, hindering development toward a fully-automated AFE workflow. The current study uses an alternative machine-learning approach to train models on a large number of solvation AFE calculations to predict the difficulty of a given perturbation a priori. Such models can replace similarity metrics to plan perturbation networks. Additionally, we generate FEP-space: given all possible perturbations in all FEP benchmarking sets, we have grafted these transformations onto a common scaffold such that we can generate a training set that encompasses the complete domain of FEP. Using this training set (n~3800) we aim to build models that outperform cutting-edge network generators.
				Ivy Zhang Memorial Sloan Kettering Cancer Center	Predicting the impact of circulating SARS-CoV-2 variants using relative binding free energy calculations: Despite recent FDA Emergency Use Authorizations of multiple therapeutics and vaccines targeting SARS-CoV-2, the rise of new viral variants (e.g. B.1.1.7, B.1.351) threatens to stunt the efficacy of these treatments. Here, we study the impact of B.1.1.7 and B.1.351 receptor binding domain (RBD) mutations N501Y, E484K, and K417N on RBD:ACE2 binding, a key protein:protein interaction for viral entry into human cells. We demonstrate that our relative binding free energy calculation software, Perses, can recapitulate the relative binding free energies of B.1.1.7 and B.1.351 RBD mutations on RBD:ACE2 binding and aim to identify new pairs of RBD mutations with compensatory effects on binding. Rapid surveillance and prediction of RBD mutations will be essential for adapting existing SARS-CoV-2 therapeutics to maintain efficacy in the presence of emergent mutations and designing novel pan-coronavirus therapeutics.
				Sofia Bariami University of Edinburgh	Implementation and Validation of the QUBE Forcefield in Sire MD Framework: Molecular mechanics (MM) force fields (FFs) are invaluable computer-aided drug design community, where rigorous free energy methods may be employed to compute the binding affinity of a candidate drug molecule to a therapeutic target. QUBE is a bespoke MM FF for small molecules and proteins that uses Quantum Mechanical (QM) calculations to derive parameters for the molecule under study. We are interested in incorporating this FF into our alchemical free energy software framework (SOMD) that supports GPU acceleration. My poster addresses how we implemented support in Sire for the QUBE FF. The implementation was tested by computing relative binding free energies for two congeneric series of non-nucleoside inhibitors of HIV-1 Reverse Transcriptase, for whom enhanced sampling has been previously conducted, using QUBE and AMBER/GAFF FFs. We observed relatively similar overall performance between the two FFs. Modifications on the bespoke parameter derivation methodologies and on Sire’s source code to include more features, can improve its accuracy.
				Adam Green Dassault Systemes BIOVIA	Application of Multi-Site Lambda Dynamics in BIOVIA Pipeline Pilot and Discovery Studio: Multi-site lambda dynamics (MSLD) is an alternative free energy method for computing relative binding and hydration free energies. MSLD has four distinct advantages over FEP: no explicit enumeration of fixed intermediate states, simulation of a competitive binding assay, simultaneous calculation of global changes to a chemical scaffold, and a novel enhanced sampling algorithm referred to as adaptive landscape flattening. The scalability of MSLD has been previously demonstrated by Vilseck et al. to screen a combinatorial library of 512 putative inhibitors of HIV Reverse Transcriptase (HIV-RT) with orders of magnitude greater efficiency over FEP with comparable accuracy. Nevertheless, this approach relied on bias potential replica exchange (BP-REX) to estimate the relative binding free energies. The implementation of MSLD in BIOVIA Pipeline Pilot/Discovery Studio does not include bias potential replica exchange; thus, we sought to test how the method implemented in Pipeline Pilot/Discovery Studio compares to the results reported by Vilseck et al. In the end, we conclude that the Pipeline Pilot/Discovery Studio approach reproduces the overall trend and shows good agreement with the experimental values for the indole inhibitors. We also discuss the possibility of extending the MSLD protocol in Pipeline Pilot/Discovery Studio to estimate pH-dependent binding free energies.
11.15am	2.12pm	7.15pm	8.15pm	POSTERS
				João Morado University of Southampton	Parameterization Made Easy With ParaMol: Thorough and accurate simulation of molecular configurational ensembles is crucial to predict theoretical static properties such as optical spectra, NMR spectra, and free energies. Owing to the computational feasibility of molecular mechanics (MM) methods, these are widely used in chemical sciences, often allowing simulation of long timescales that permit ergodic sampling to be achieved or, at least, approached. In this context, force fields (FF) are commonly employed to describe the potential energy function of systems of interest and to simulate their dynamical behaviour. Despite FFs’ low computational cost, their accuracy is frequently hindered by functional form constraints and poor transferability of parameterization. Among the several functional forms proposed thus far, the class I additive potential energy function is employed in the majority of atomistic biomolecular simulations, for which accurate FF parameters for biomolecules like proteins or DNA are available. Nevertheless, reliable parameters are usually unavailable for novel molecules such as drug candidates, as these may involve functional groups and interactions that are particularly challenging and system specific. In this poster, we present the ParaMol [1], software that has a special focus on the parameterization of bonded and nonbonded terms of class I FFs by fitting to ab initio data. We describe the theory underlying ParaMol’s parameterization philosophy, as well as its several features, viz. the capability of performing self-consistent parameterization, dihedral scans, RESP fitting, among other available tasks. Additionally, we illustrate the software’s capabilities with application examples and report the best practices to be applied alongside each parameterization recipe. Owing to ParaMol’s capabilities and high efficiency, we propose that this software can be introduced as a routine step in the protocol normally employed to parameterize druglike molecules for MM simulations. [1] Morado, J.; Mortenson, P. N.; Verdonk, M. L.; Ward, R. A.; Essex, J. W.; Skylaris, C.-K. ParaMol: A Package for Automatic Parameterization of Molecular Mechanics Force Fields. J. Chem. Inf. Model. 2021, 61 (4), 2026–2047
				Martin Reinhardt Max Planck Institute for Biophysical Chemistry	Calculating Free Energy Differences Through Variationally-derived Intermediates: Alchemical methods commonly require morphing between the Hamiltonians of the involved molecular end states, typically in terms of a linear interpolation. However, linear transformations and soft-core variants thereof are still a very special case amongst all possible transformations. Here, we generalized this approach and derived, under the assumption of uncorrelated sampling points, the set of discrete intermediates with optimal sampling accuracy - hence labelled the Variationally-derived Intermediates (VI) - using Free Energy Perturbation and the Bennett Acceptance Ratio method as estimators. The VI are optimal for any number of intermediate states and sample points. Interestingly, these optimal intermediates are coupled through a system of equations and do not require a user choice of a lambda variable. We present applications to the case of solvation free energies, yielding improved accuracies for decoupling electrostatic interactions, and similar ones for Lennard-Jones interactions, compared to state-of-the-art soft-core transformations.
				Nitin Singh Indian Institute of Technology, Gandhinagar	Understanding the Helical Stability of Charged Peptides: α-helices play an essential role in the tertiary and quaternary folding of a protein and are vital for protein functioning. Helical peptides have found their applications in membrane transport, vaccine development, and therapeutics. Their application is limited because of the reduced solubility of the hydrophobic residue helices and the low stability of the helices with charged amino acids in the aqueous solution. This study aims to bridge this gap by designing water-soluble helical peptides by controlling the charge density and the amino acid sequence. In this study, we used leucine (hydrophobic) and lysine (charged) amino acid residues to design proteins with considerably higher stability than fully charged polypeptides. Results from Molecular Dynamics Simulations show that charge density plays a central role in tuning the helical stability. At a fixed charge density, the sequence pattern has a minor influence. We believe this study could help the scientific community involved in the de novo design of protein sequences.
				Salomé Rieder ETH Zurich	Using GAFF topologies and RE-EDS to calculate relative hydration free energies in GROMOS: Free-energy differences between pairs of end states can be estimated based on molecular dynamics (MD) simulations using standard methods such as thermodynamic integration (TI). Replica-exchange enveloping distribution sampling (RE-EDS), on the other hand, allows the sampling of multiple states in a single simulation. In this work, relative hydration free energies are calculated for a series of benzene derivatives using the RE-EDS approach in GROMOS. GAFF topologies are converted to a GROMOS-compatible format with the novel amber2gromos program. The results obtained with RE-EDS are compared to the experimental and simulated values from the FreeSolv database. In addition, the estimated free-energy differences in water and in vacuum are compared to TI simulations carried out with GROMACS. The results show that the hydration free energies calculated with RE-EDS for multiple ligands are in good agreement with both the experimental data and pairwise TI values. This work serves as a validation that GAFF topologies can be used with the GROMOS simulation package and the RE-EDS approach.
				Drazen Petrov BOKU, University of Natural Resources and Life Sciences, Vienna, Austria	Perturbation free-energy toolkit: automated alchemical topology builder and adaptive simulation update scheme: Free-energy calculations play an important role in the application of computational chemistry to a range of fields, including protein biochemistry, rational drug design or material science. This study aims at addressing two challenges related with perturbation approaches: 1) the definition of the perturbation path, i.e., alchemical changes leading to the transformation of one molecule to the other, and 2) determining the amount and distribution of simulation along the path to reach desired convergence. An automatic perturbation builder based on a graph matching algorithm is developed, that can identify the maximum common substructure of two molecules and provide the perturbation topologies suitable for free-energy calculations using GROMOS and GROMACS simulation packages. Moreover, it was used to calculate the changes in free energy of a set of post-translational modifications and analyze their convergence behavior in various simulation scenarios. Importantly, this toolkit is made available online as an open-source python package (https://github.com/drazen-petrov/SMArt).
				David Zierke Leibniz-Forschungsinstitut für Molekulare Pharmakologie (FMP Berlin)	Prediction of Factor Xa - Inhibitor Binding Affinities using Multisite Lambda Dynamics: The assessment of changes in protein-ligand binding affinities in response to variation of substituents is of great importance in lead structure optimization, but still limited to small compound sets due to high computational cost. Recent implementation of the Multisite Lambda Dynamics (MSLD) workflow in BIOVIA Discovery Studio makes it possible to explore substantially larger molecule sets with variation on different sites. We employed this approach and showed here some preliminary results for relative binding free energy calculations of factor Xa - inhibitor complexes. Using the automated MSLD workflow for a set of high-affinity ligands from different series, we obtained reasonable predictions on multi-site variations with muss less computational cost compared to existing alchemical free energy calculation methods. Our results thus encourage the application of MSLD in a structure-based drug design process.

POSTERS Thursday

Some presentations can be viewed ahead of the meeting here

PDT	ET	UK	CET	Tuesday 15th June 2021	Abstract
10.50am	1.50pm	6.50pm	7.50pm	POSTERS
				Mateusz K. Bieniek University of Newcastle	TIES 20: Relative Binding Free Energy with a Flexible Superimposition Algorithm and Partial Ring Morphing: The alchemical approach to Relative Binding Free Energy calculations requires a high quality mapping of atoms across compared ligands. Our new software TIES 20 uses a flexible recursive joint-traversal algorithm for the generation of the Maximum Common Substructure. On average, by enabling partial ring morphing, smaller alchemical regions were computed. We evaluated the software with NAMD 2 using the dual topology approach. We show that there is a clear relationship between the size of the alchemical region and the uncertainty in the predicted binding energies. We further show that AM1-BCC modifies the charges closer to the site of the mutation, decreasing the size of the alchemical region substantially, allowing for easier convergence. Our results demonstrate that having 5 replicas per lambda window significantly improves reproducibility of the binding energies in comparison to just one replica, but even so the reproducibility highly depends on the complexity of the system.
				Yunhui Ge University of California, Irvine	Enhancing sampling of water rearrangements on ligand binding: A comparison of techniques: Water often plays a key role in protein structure, molecular recognition, and mediating protein-ligand interactions. Thus, free energy calculations must adequately sample water motions, which often proves challenging in typical MD simulation timescales. Thus, the accuracy of methods relying on MD simulations ends up limited by slow water sampling. Particularly, as a ligand is removed or modified, bulk water may not have time to fill or rearrange in the binding site. In this work, we focus on several molecular dynamics (MD) simulation-based methods attempting to help address water motions and occupancies: BLUES, using nonequilibrium candidate Monte Carlo (NCMC); \textit{grand}, using grand canonical Monte Carlo (GCMC); and normal MD. We assess the accuracy and efficiency of these methods in sampling water motions. We selected a broad range of systems with varying numbers of waters in the binding site, as well as those where water occupancy is coupled to the identity or binding mode of the ligand. We analyzed water motions and occupancies using both clustering of trajectories and direct analysis of electron density maps. Our results suggest both BLUES and \textit{grand} enhance water sampling relative to normal MD and \textit{grand} is more robust than BLUES, but also that water sampling remains a major challenge for all of the methods tested. The lessons we learned for these methods and systems are discussed.
				Esteban Vohringer Universidad de Concepcion, Chile	Environment Specific D-MBIS Atomic Charges Improve Binding Free Energy Predictions of SAMPL5 Host-Guest Systems: We will present a computational pipeline which generates atomic charges for guest molecules in the bound and unbound state of host-guest systems and calculates the average polarization energy in each molecular environment with the QM/MM methodology. Using a validated protocol in the Yank software, we employ each charge set and the polarization energy in alchemical free energy calculations to obtain binding free energies in one day on a standard GPU. Our results improve binding free energies obtained with the AM1-BCC or RESP charges and rank the guests by affinity correctly in most of the cases.
				Ernest Williams Memorial University of Newfoundland	Modelling the Binding Free Energy of Peptidomimetic inhibitors of SARS CoV-2 Main Protease
				Germano Heinzelmann Universidade Federal de Santa Catarina (UFSC)	Automated tools for absolute binding free energy calculations: We present here the 2.0 version of the Binding Affinity Tool (BAT.py), and the Guest-HOst Affinity Tool (GHOAT.py), two free and open-source programs that compute absolute binding free energies (ABFE) in a fully automated way. They use the standard decoupling and recoupling (SDR) method with restraints, which can be applied to neutral and charged ligands without the need for analytical corrections. BAT.py is designed for both docking refinement and ligand-protein affinity evaluation, combining automation with the exceptionally low cost of simulations using AMBER20 on GPUs. BAT opens the door for ABFE to be applied in a high-throughput regime, aiding in the search for new high-affinity molecular fragments and scaffolds. GHOAT.py is applied to host-guest systems, used as simple models for parameter testing and evaluation. GHOAT requires minimal input for a new host, and is able to calculate the affinity of any guest that binds to it without any manual steps.
				Krystel El Hage University Paris-Saclay	Inhibiting RNA : Protein Interactions using an Integrative Computational and Experimental Approach: Application to YB1: While targeting Protein:Protein interactions has served as basis for the development of new drugs, RNA:Protein interactions, which are as important in human pathologies, notably cancer, is very promessing but remains largely unexploited. One of the major mRNA binding proteins is YB-1, a master regulator of translation in cancer cells. It has been recently considered as a therapeutic target for the treatment of cancer and drug-resistant cancer. However, no small molecules with high affinity and specificity against YB- 1 have been proposed so far along with the structural data essential to interrogate their relevance. The conception of drug candidates is a very delicate procedure that requires a prior understanding of the translation regulation systems at the atomistic level using a high level of accuracy and state-of-the-art techniques. And connecting in silico data to structural data from diverse experimental resources is mandatory to provide a resolved picture of the underlying mechanisms. Hence, combining both advanced computational and experimental techniques will help integrate chemical, structural and cellular data for the development of small molecules that target RNA:YB-1 interactions. This is done by developing an innovative approach that would integrate a) molecular modeling data (Drug design, Molecular Dynamics and Free Energy Simulations using a sufficiently accurate computational model that is computationally efficient), b) NMR Spectroscopy data, together with c) an experimental validation in cells with a new HCS technology, “MT Bench”, that quantifies RNA:protein interactions at the single cell level. The major advantage is providing cellular and structural data to feed, with little delay, the computational approach to propose efficient and specific ligands that target translation regulation in vitro and in cancer cells. This approach allowed us, for now, to identify several promising active compounds in the sub-micromolar range. We will focus, in a next step, on rationally optimizing them to help propose new anti-cancer drugs able to overcome drug resistance by targeting YB-1 with a higher affinity and a larger selectivity.
				Himanshu Goel University of Maryland, Baltimore	Capturing Water Networks During Ligand Binding with the Site-Identification by Ligand Competitive Saturation Approach: Water molecules impact interactions between proteins and their ligands, making a significant contribution to ligand binding orientation and affinity. Water molecules can significantly affect the energetics of ligand binding in a favorable or unfavorable manner associated, in part, with the energetic penalty for displacing waters that occupy a binding site. The challenge is to calculate water mediated interactions and their energetics between the ligand and the protein. In this work, an extension of the Site Identification by Ligand Competitive Saturation-Monte Carlo (SILCS-MC) docking approach is presented to determine the positioning of the water molecules in the binding site and leads to improve the correlation and rank-ordering scores. This also provide an information about the energetic contribution of water to ligand binding. We employ this methodology on variety of protein targets where water mediated interactions between the protein and ligands plays an important role. The efficacy of this approach is reflected in rank-ordering ligand affinities, binding affinities predictions, and structure orientation of the ligands. The presented approach offers new possibilities in revealing water networks and their contributions to the binding affinity of a ligand to a protein.

Introduction

This workshop will focus on the current state of free energy techniques.

Our goal in this workshop is to bring together experts from pharma and supporting industries, as well as academia, in an intense and focused workshop to identify challenges and help chart the path forward. We are particularly interested in hearing about use cases, pitfalls and their solutions. We also firmly believe we can learn a great deal from failure, so we hope participants will go beyond just highlighting success stories to provide more detailed insight into successes and failures.

Dates and Location

The conference will be held 6th - 8th May 2020, finishing at 3pm on Friday 8th, at the Novartis Institute for BioMedical Research, Cambridge, MA.

MEETING HAS BEEN POSTPONED UNTIL 2022

The following social arrangements are in place for the evenings of the conference.

Tuesday evening - a reception will be held at Relay Therapeutics, 399 Binney St.

Wednesday evening - a poster session will be held at the conference venue, Novartis.

Thursday evening - Schrödinger will sponsor an evening session at a nearby venue - TBD.

Virtual Conference - November 2020

11-13th November 2020

View the conference talk recordings HERE

PDT	ET	UK	CET	Wednesday 11th November 2020
8.30am	11.30am	4.30pm	5.30pm	Introduction
				Session 1 Moderator : Camilo Velez-Vega
8.45am	11.45am	4.45pm	5.45pm	Francesca Deflorian, Sosei Heptares, Prediction of ligand binding affinity with Free Energy Perturbation for GPCRs
9.10am	12.10pm	5.10pm	6.10pm	Mark Mackey, Cresset, Automated Assessment of Binding Affinity via Alchemical Free Energy Calculations --- recorded
9.35am	12.35pm	5.35pm	6.35pm	Aysegul Ozen, Blueprint Medicines, FEP Impact on Early and Late-Stage Programs across Blueprint Medicines Portfolio --- recorded
10am	1pm	6pm	7pm	break
				Session 2 Moderator : Sereina Riniker
10.30am	1.30pm	6.30pm	7.30pm	Alex Dickson, Michigan State University, Binding free energies from weighted ensemble path sampling --- recorded
10.55am	1.55pm	6.55pm	7.55pm	Ido Ben-Shalom, University of California, San Diego, Exchanging Buried Water in Protein-Ligand Free Energy Calculations by Monte Carlo/Molecular Dynamics Sampling --- recorded
11.20am	2.20pm	7.20pm	8.20pm	Maximilian Ebert, CCG, Soft-core potential optimization to increase accuracy and stability in Free Energy Calculations using AMBER Thermodynamic Integration in MOE --- recorded

PDT	ET	UK	CET	Thursday 12th November 2020
8.30am	11.30am	4.30pm	5.30pm	Introduction
				Session 1 Moderator : Zoe Cournia
8.45am	11.45am	4.45pm	5.45pm	Daria Kokh, Heidelberg Institute for Theoretical Studies, Exploring ligand unbinding kinetics using random acceleration: what can we learn? --- recorded
9.10am	12.10pm	5.10pm	6.10pm	Hugo Gutierrez de Teran, University of Uppsala, Q-FEP: Versatile, robust and automated FEP protocols to integrate SAR and mutagenesis data on ligand design --- recorded
9.35am	12.35pm	5.35pm	6.35pm	Heather Carlson, University of Michigan, MixMD: Mapping Protein Surfaces
10am	1pm	6pm	7pm	break
				Session 2 Moderator : Kira Armacost
10.30am	1.30pm	6.30pm	7.30pm	Jay Ponder, Washington University in St. Louis, AMOEBA Host-Guest Binding Free Energies for the SAMPL7 Challenge --- recorded
10.55am	1.55pm	6.55pm	7.55pm	Prabhu Raman, BIOVIA, Efficient relative binding free energy calculations using Lambda Dynamics --- recorded
11.20am	2.20pm	7.20pm	8.20pm

PDT	ET	UK	CET	Friday 13th November 2020
8.30am	11.30am	4.30pm	5.30pm	Introduction
				Session 1 Moderator : Jonah Vilseck
8.45am	11.45am	4.45 -5.10	5.45 - 6.10	Christina Schindler, Merck KGaA, Evaluation of free energy calculations on the Merck FEP benchmark datasets
9.10am	12.10pm	5.10 - 5.35	6.10 - 6.35	Lingle Wang, Schrodinger, Improving the Accuracy of Free Energy Calculations via Enhanced Sampling of Water Relaxation, Ionization Equilibrium and Protein Conformational Changes --- recorded
9.35am	12.35pm	5.35pm	6.35pm	Darrin York, Rutgers University, Emerging methods for accurate protein-ligand binding free energy prediction in AMBER --- recorded
10am	1pm	6pm	7pm	break
				Session 2 Moderator : Michael Gilson
10.30am	1.30pm	6.30pm	7.30pm	Lance Westerhoff, QuantumBio Inc., MoveableType: A fast, accurate method for binding free energy prediction and simulation --- recorded
10.55am	1.55pm	6.55pm	7.55pm	Chia-en Chang, University California Riverside, Tuning ligand binding kinetics using transient states and free energy barriers --- recorded
11.20am	2.20pm	7.20pm	8.20pm	Closing remarks

Accommodation

Discount rates may be available for some hotels in the area. Novartis suggests that attendants individually call CWT at 1-800-929-2231 and ask for the “Novartis discount rate” at the following hotels (mentioning their attendance to a workshop at NIBR):

Le Meridian (20 Sidney Street, Cambridge MA)

Hyatt Regency (575 Memorial Drive, Cambridge MA)

Kimpton Marlowe Hotel (25 Edwin H Land Blvd, Boston MA)

Cambridge Marriott (120 Broadway Six Cambridge Center, Cambridge MA)

Novartis is close to the Central stop on the red line, as well as the number 1 bus route, and there may be more affordable accommodation options using these transport links.

Registration

Registration will cost $100 for admission to the 2.5 day conference, which will cover both breakfast and lunch, as well as refreshments in the evening sessions. We will also be accepting submitted abstracts for oral and poster presentations at this time. The deadline for oral presentations is December 16th 2019, while poster presentations is March 27th 2020 DUE TO POSTPONEMENT: June 26th 2020. Submitting an abstract alone does not guarantee you a place, please make sure to register too.

• Registration is full! To join the waitlist please contact Hannah Bruce Macdonald (hannah.brucemacdonald@choderalab.org). If you are no longer planning to attend, please let us know and we can refund your fee.
• Abstract submission is here

Workshop Themes

As this is the 10th birthday of the first 2010 free energy workshop, we will be holding a session celebrating the last 10 years - how far have we come, and where do we still need to go?

Other sessions will be held to discuss the current software that is currently available to the field - both commercial packages, and open-source academic efforts.

Discussions about the novel methods and algorithms that are in development and where these can be best applied will be held.

Results of free energy methods will be presented - both the successes and the limitations - with the idea of addressing what the community would like to see at the following meeting.

Finally, half a day of talks will be dedicated to the intersection of free energy methods with machine learning.

JCIM Special Issue

JCIM is aiming to publish a special issue on Novel Directions in Free Energy Methods and Applications for publication in the Fall of 2020. The submission deadline is June 1st, 2020. This is spearheaded by Kira Armacost, Sereina Riniker and Zoe Cournia. For more information, please read here.

Sponsors, Social Media and Streaming

Sponsors

We would like to thank the following sponsors:

We are currently looking for sponsorship for the event. If your organisation would be interested, please contact any of the organisers.

Particular thanks to the Boston Area Group for Informatics and Modeling (BAGIM)for helping with the financial organisation of this meeting.

Social Media

Twitter hashtags: #alchemy2020
Slack: https://alchemistry.slack.com

Streaming

We will be streaming the conference, and uploading talks after the conference, however the recording of any talk is at the discretion of the speaker, their employer and collaborators.

Schedule

Talks will be held from 9am to 6pm (Wednesday and Thursday) or 3pm (Friday) at Novartis.

Keynote speaker

• Herman van Vlijmen, Janssen

Confirmed speakers

• Robert Abel, Schrödinger
• Christopher Bayly, OpenEye Scientific
• Ido Y. Ben-Shalom, University of California, San Diego
• Phil Biggin, University of Oxford
• Heather Carlson, University of Michigan
• Chia-en Chang, University California Riverside
• Chris Chipot, University of Illinois
• Lucy Colwell, University of Cambridge
• Francesca Deflorian, Sosei Heptares
• Alex Dickson, Michigan State University
• Jonathan Essex, University of Southampton
• Gianni de Faabrits, Acellera
• Andrey Frolov, AstraZeneca
• Niels Hansen, University of Stuttgart
• William Jorgensen, Yale University
• Dariah Kokh, Heidelberg Institute for Theoretical Studies
• Wonpil Im, Lehigh University
• Yun Lyna Luo, Western University of Health Sciences
• Mark Mackey, Cresset
• Katarina Meier, Bayer
• David Mobley, University California, Irvine
• Julien Michel, University of Edinburgh
• David Minh, Illinois Institute of Technology
• Aysegul Ozen, Blueprint Medicines
• Jay Ponder, Washington University in St. Louis
• Prabhu Raman, BIOVIA
• Sereina Riniker, ETH Zürich
• Christina Schindler, Merck KGaA
• Michael Schnieders, University of Iowa
• Woody Sherman, Silicon Therapeutics
• Michael Shirts, University of Colorado, Boulder
• Hugo Gutierrez de Teran, University of Uppsala
• David Thompson, CCG
• Jonah Vilseck, Indiana University School of Medicine
• Lingle Wang, Schrödinger
• Lance Westerhoff, QuantumBio Inc.
• Xin Cindy Yan, Merck & Co., Inc.
• Joy Yang, Pfizer
• Darrin York, Rutgers University
• Yutong Zhao, Relay Therapeutics

Organizers

• Kira Armacost, GlaxoSmithKline Pharmaceuticals (kira.x.armacost@gsk.com)
• Hannah Bruce Macdonald, Memorial-Sloan Kettering Cancer Center (hannah.brucemacdonald@choderalab.org)
• Jonah Vilseck, Indiana University School of Medicine (jvilseck@iu.edu)
• Zoe Cournia, Biomedical Research Foundation Academy of Athens (zcournia@bioacademy.gr)
• Michael Gilson, UC San Diego (mgilson@ucsd.edu)
• Camilo Velez-Vega, Novartis Institutes for BioMedical Research (camilo.velez-vega@novartis.com)
• Sereina Riniker, ETH Zürich (sriniker@ethz.ch)
• Anthony Arvanites, Eli Lilly and BAGIM (aarvanites@bagim.org)

We also thank current advisors and previous organizers:

• Brian McClain, Vertex (brian_mcclain@vrtx.com)
• Vijay Pande, Stanford University (pande@stanford.edu)
• Michael Shirts, University of Colorado Boulder (michael.shirts@colorado.edu)
• Camilo Velez-Vega, Novartis Institutes for BioMedical Research (camilo.velez-vega@novartis.com)
• John Chodera, Memorial-Sloan Kettering Cancer Center (john.chodera@choderalab.org)
• Greg Bowman, Washington University in St. Louis (bowman@biochem.wustl.edu)
• Callum Dickson, Novartis (callum.dickson@novartis.com)
• Jose Duca, Novartis (jose.duca@novartis.com)
• Viktor Hornak, Novartis (viktor.hornak@novartis.com)
• John Manchester, Novartis (john.manchester@novartis.com)
• Antonia Mey, University of Edinburgh (antonia.mey@ed.ac.uk)
• David Mobley, University of California at Irvine (dmobley@uci.edu)
• Michael Schnieders, The University of Iowa (michael-schnieders@uiowa.edu)
• Jana Shen, The University of Maryland (Jana.Shen@rx.umaryland.edu)
• Woody Sherman, Silicon Therapeutics (woody@silicontx.com)

Conference Code of Conduct

All attendees, speakers, sponsors and volunteers at our conference are required to agree with the following code of conduct. Organisers will enforce this code throughout the event. We expect cooperation from all participants to help ensure a safe environment for everybody.

Need Help? Please contact the organising committee either in person or via email. Members of the organising committee are listed above.

The Quick Version

Our conference is dedicated to providing a harassment-free conference experience for everyone, regardless of gender, gender identity and expression, age, sexual orientation, disability, physical appearance, body size, race, ethnicity, religion (or lack thereof), or technology choices. We do not tolerate harassment of conference participants in any form. Sexual language and imagery is not appropriate for any conference venue, including talks, workshops, parties, Twitter and other online media. Conference participants violating these rules may be sanctioned or expelled from the conference without a refund at the discretion of the conference organisers.

The Less Quick Version

Harassment includes offensive verbal comments related to gender, gender identity and expression, age, sexual orientation, disability, physical appearance, body size, race, ethnicity, religion, technology choices, sexual images in public spaces, deliberate intimidation, stalking, following, harassing photography or recording, sustained disruption of talks or other events, inappropriate physical contact, and unwelcome sexual attention.

Participants asked to stop any harassing behavior are expected to comply immediately.

Sponsors are also subject to the anti-harassment policy. In particular, sponsors should not use sexualised images, activities, or other material. Booth staff (including volunteers) should not use sexualised clothing/uniforms/costumes, or otherwise create a sexualised environment.

If a participant engages in harassing behavior, the conference organisers may take any action they deem appropriate, including warning the offender or expulsion from the conference with no refund.

If you are being harassed, notice that someone else is being harassed, or have any other concerns, please contact a member of the organising committee immediately.

A member of the organising committee will be happy to help participants contact local law enforcement, provide escorts, or otherwise assist those experiencing harassment to feel safe for the duration of the conference. We value your attendance.

We expect participants to follow these rules at conference and workshop venues and conference-related social events.

The code of conduct text was taken from https://github.com/confcodeofconduct/confcodeofconduct.com in a slightly adapted form.