Method development
Small Molecule Modulation of Intrinsically Disordered Proteins Using Molecular Dynamics Simulations.
WU tags:
Description
The extreme dynamic behavior of intrinsically disordered proteins hinders the development of drug-like compounds capable of modulating them. There are several examples of small molecules that specifically interact with disordered peptides. However, their mechanisms of action are still not well understood. Here, we use extensive molecular dynamics simulations combined with adaptive sampling algorithms to perform free ligand binding studies in the context of intrinsically disordered proteins. We tested this approach in the system composed by the D2 sub-domain of the disordered protein
Publications
- Herrera-Nieto P, Pérez A, De Fabritiis G, Small Molecule Modulation of Intrinsically Disordered Proteins Using Molecular Dynamics Simulations. Journal of chemical information and modeling 2020. doi:10.1021/acs.jcim.0c00381
| Badge | Rank | Name | Credit |
|---|---|---|---|
 | 1 | Rion Family | 2,436,730,050.00 |
| 2 | grcpool.com-2 | 1,829,466,625.00 |
| 3 | grcpool.com | 1,606,947,375.00 |
| 4 | Aurum | 1,522,558,100.00 |
| 5 | jjch | 1,509,867,350.00 |
| 6 | Retvari Zoltan | 1,432,138,525.00 |
| 7 | grcpool.com-3 | 1,370,851,650.00 |
| 8 | Viktor Svantner | 1,047,191,550.00 |
| 9 | PappaLitto | 904,308,225.00 |
| 10 | Zalster | 853,586,300.00 |
GPCRmd uncovers the dynamics of the 3D-GPCRome.
WU tags:
Description
G-protein-coupled receptors (GPCRs) are involved in numerous physiological processes and are the most frequent targets of approved drugs. The explosion in the number of new three-dimensional (3D) molecular structures of GPCRs (3D-GPCRome) over the last decade has greatly advanced the mechanistic understanding and drug design opportunities for this protein family. Molecular dynamics (MD) simulations have become a widely established technique for exploring the conformational landscape of proteins at an atomic level. However, the analysis and visualization of MD simulations require efficient storage resources and specialized software. Here we present GPCRmd (http://gpcrmd.org/), an online platform that incorporates web-based visualization capabilities as well as a comprehensive and user-friendly analysis toolbox that allows scientists from different disciplines to visualize, analyze and share GPCR MD data. GPCRmd originates from a community-driven effort to create an open, interactive and standardized database of GPCR MD simulations.
Publications
- Rodríguez-Espigares I, Torrens-Fontanals M, Tiemann JKS, Aranda-García D, Ramírez-Anguita JM, Stepniewski TM, Worp N, Varela-Rial A, Morales-Pastor A, Medel-Lacruz B, Pándy-Szekeres G, Mayol E, Giorgino T, Carlsson J, Deupi X, Filipek S, Filizola M, Gómez-Tamayo JC, Gonzalez A, Gutiérrez-de-Terán H, Jiménez-Rosés M, Jespers W, Kapla J, Khelashvili G, Kolb P, Latek D, Marti-Solano M, Matricon P, Matsoukas MT, Miszta P, Olivella M, Perez-Benito L, Provasi D, Ríos S, R Torrecillas I, Sallander J, Sztyler A, Vasile S, Weinstein H, Zachariae U, Hildebrand PW, De Fabritiis G, Sanz F, Gloriam DE, Cordomi A, Guixà-González R, Selent J, GPCRmd uncovers the dynamics of the 3D-GPCRome. Nature methods 2020. doi:10.1038/s41592-020-0884-y
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | Retvari Zoltan | 107,681,600.00 |
| 2 | Rion Family | 96,720,000.00 |
| 3 | grcpool.com | 83,904,600.00 |
| 4 | grcpool.com-2 | 74,393,800.00 |
| 5 | PappaLitto | 57,790,200.00 |
| 6 | RaymondFO* | 55,775,200.00 |
| 7 | jjch | 53,115,400.00 |
| 8 | grcpool.com-3 | 51,664,600.00 |
| 9 | Stoneageman | 50,858,600.00 |
| 10 | vanos0512 | 32,240,000.00 |
PlayMolecule CrypticScout: Predicting Protein Cryptic Sites Using Mixed-Solvent Molecular Simulations.
WU tags:
Description
Cryptic pockets are protein cavities that remain hidden in resolved apo structures and generally require the presence of a co-crystallized ligand to become visible. Finding new cryptic pockets is crucial for structure-based drug discovery to identify new ways of modulating protein activity and thus expand the druggable space. We present here a new method and associated web application leveraging mixed-solvent molecular dynamics (MD) simulations using benzene as a hydrophobic probe to detect cryptic pockets. Our all-atom MD-based workflow was systematically tested on 18 different systems and 5 additional kinases and represents the largest validation study of this kind. CrypticScout identifies benzene probe binding hotspots on a protein surface by mapping probe occupancy, residence time, and the benzene occupancy reweighed by the residence time. The method is presented to the scientific community in a web application available via www.playmolecule.org using a distributed computing infrastructure to perform the simulations.
Publications
- Martinez-Rosell G, Lovera S, Sands ZA, De Fabritiis G, PlayMolecule CrypticScout: Predicting Protein Cryptic Sites Using Mixed-Solvent Molecular Simulations. Journal of chemical information and modeling 2020. doi:10.1021/acs.jcim.9b01209
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | Rion Family | 4,639,746.00 |
| 2 | Retvari Zoltan | 4,450,949.00 |
| 3 | grcpool.com | 3,535,477.00 |
| 4 | Zalster | 2,362,054.00 |
| 5 | grcpool.com-2 | 1,986,559.00 |
| 6 | Stoneageman | 1,969,784.00 |
| 7 | grcpool.com-3 | 1,945,416.00 |
 | 8 | jjch | 1,834,758.00 |
| 9 | Viktor Svantner | 1,796,967.00 |
| 10 | Erich56 | 1,755,579.00 |
Machine Learning of Coarse-Grained Molecular Dynamics Force Fields.
WU tags:
Description
Atomistic or ab initio molecular dynamics simulations are widely used to predict thermodynamics and kinetics and relate them to molecular structure. A common approach to go beyond the time- and length-scales accessible with such computationally expensive simulations is the definition of coarse-grained molecular models. Existing coarse-graining approaches define an effective interaction potential to match defined properties of high-resolution models or experimental data. In this paper, we reformulate coarse-graining as a supervised machine learning problem. We use statistical learning theory to decompose the coarse-graining error and cross-validation to select and compare the performance of different models. We introduce CGnets, a deep learning approach, that learns coarse-grained free energy functions and can be trained by a force-matching scheme. CGnets maintain all physically relevant invariances and allow one to incorporate prior physics knowledge to avoid sampling of unphysical structures. We show that CGnets can capture all-atom explicit-solvent free energy surfaces with models using only a few coarse-grained beads and no solvent, while classical coarse-graining methods fail to capture crucial features of the free energy surface. Thus, CGnets are able to capture multibody terms that emerge from the dimensionality reduction.
Publications
- Wang J, Olsson S, Wehmeyer C, Pérez A, Charron NE, de Fabritiis G, Noé F, Clementi C, Machine Learning of Coarse-Grained Molecular Dynamics Force Fields. ACS central science 2019. doi:10.1021/acscentsci.8b00913
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | Aurum | 1,389,750.00 |
| 2 | grcpool.com | 1,332,525.00 |
| 3 | grcpool.com-2 | 1,188,100.00 |
| 4 | grcpool.com-3 | 1,073,650.00 |
| 5 | Zalster | 915,600.00 |
| 6 | Stoneageman | 700,325.00 |
| 7 | SB152UcAvM6XzgQrBUDySTf4EhiztpoKAc | 621,300.00 |
| 8 | Retvari Zoltan | 539,550.00 |
| 9 | JStateson | 506,850.00 |
| 10 | mm67 | 490,500.00 |
Molecular Simulation-Driven Fragment Screening for the Discovery of New CXCL12 Inhibitors
WU tags: *CXCL12_LIG*
Description
![[Experiment image]](img/science/methods/toc_cxcl12.png)
Fragment-based drug discovery (FBDD) has been proposed as an alternative to classical high-throughput screening techniques, in which milions of compounds are screened to find potential drugs. Instead, in FBDD, the smaller size of the compounds allows us to reduce screening libraries to only hundreds of compounds. In this work we apply molecular dynamics (MD) and a Markov State Model (MSM) framework to screen a library of 129 compounds against the protein CXCL12, a chemokine related to many diseases such as cancer methastasis. We are able to identify up to 8 fragments with millimolar affinity that bind to two pockets of the chemokine, named sY7 and H1S68. This work paves the way for the introduction of MD-based techniques in mainstream drug discovery pipelines.
Publications
- Martinez-Rosell G, Harvey MJ, De Fabritiis G. Molecular Simulation-Driven Fragment Screening for the Discovery of New CXCL12 Inhibitors. J Chem Inf Model. 2018 Feb 26. doi: 10.1021/acs.jcim.7b00625
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | Stoneageman | 1,729,897,300.00 |
| 2 | RaymondFO* | 1,505,662,625.00 |
| 3 | Retvari Zoltan | 1,139,283,575.00 |
| 4 | Rion Family | 950,791,400.00 |
| 5 | jjch | 941,778,375.00 |
| 6 | HA-SOFT, s.r.o. | 907,387,500.00 |
| 7 | Acey Pilot | 808,675,025.00 |
| 8 | caffeineyellow5 | 796,997,950.00 |
| 9 | Beyond | 713,431,075.00 |
| 10 | neilp62 | 656,984,750.00 |
Complete protein-protein association kinetics in atomic detail revealed by molecular dynamics simulations and Markov modelling.
WU tags:
Description
Protein-protein association is fundamental to many life processes. However, a microscopic model describing the structures and kinetics during association and dissociation is lacking on account of the long lifetimes of associated states, which have prevented efficient sampling by direct molecular dynamics (MD) simulations. Here we demonstrate protein-protein association and dissociation in atomistic resolution for the ribonuclease barnase and its inhibitor barstar by combining adaptive high-throughput MD simulations and hidden Markov modelling. The model reveals experimentally consistent intermediate structures, energetics and kinetics on timescales from microseconds to hours. A variety of flexibly attached intermediates and misbound states funnel down to a transition state and a native basin consisting of the loosely bound near-native state and the tightly bound crystallographic state. These results offer a deeper level of insight into macromolecular recognition and our approach opens the door for understanding and manipulating a wide range of macromolecular association processes.
Publications
- Plattner N, Doerr S, De Fabritiis G, Noé F, Complete protein-protein association kinetics in atomic detail revealed by molecular dynamics simulations and Markov modelling. Nature chemistry 2017. doi:10.1038/nchem.2785
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | Syracuse University | 6,079,726,000.00 |
| 2 | RaymondFO* | 3,788,924,550.00 |
| 3 | Retvari Zoltan | 3,513,857,000.00 |
| 4 | jjch | 2,618,193,350.00 |
| 5 | Rion Family | 2,143,561,300.00 |
| 6 | Beyond | 1,864,099,850.00 |
| 7 | Greger | 1,847,771,600.00 |
| 8 | Bob Deller | 1,696,910,250.00 |
| 9 | Stoneageman | 1,584,619,650.00 |
| 10 | AND | 1,450,440,550.00 |
High-Throughput Automated Preparation and Simulation of Membrane Proteins with HTMD.
WU tags:
Description
HTMD is a programmable scientific platform intended to facilitate simulation-based research in molecular systems. This paper presents the functionalities of HTMD for the preparation of a molecular dynamics simulation starting from PDB structures, building the system using well-known force fields, and applying standardized protocols for running the simulations. We demonstrate the framework's flexibility for high-throughput molecular simulations by applying a preparation, building, and simulation protocol with multiple force-fields on all of the seven hundred eukaryotic membrane proteins resolved to-date from the orientation of proteins in membranes (OPM) database. All of the systems are available on www.playmolecule.org .
Publications
- Doerr S, Giorgino T, Martínez-Rosell G, Damas JM, De Fabritiis G, High-Throughput Automated Preparation and Simulation of Membrane Proteins with HTMD. Journal of chemical theory and computation 2017. doi:10.1021/acs.jctc.7b00480
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | Syracuse University | 366,198,330.00 |
| 2 | Retvari Zoltan | 174,941,078.00 |
| 3 | RaymondFO* | 149,034,581.00 |
| 4 | jjch | 134,256,193.00 |
| 5 | Bob Deller | 127,494,569.00 |
| 6 | Rion Family | 122,850,755.00 |
| 7 | Beyond | 110,804,013.00 |
| 8 | Greger | 109,300,080.00 |
| 9 | Bedrich Hajek | 84,518,514.00 |
| 10 | Erich56 | 72,719,027.00 |
Insights from Fragment Hit Binding Assays by Molecular Simulations
WU tags: *XA*, (10|18|27|29|31)x*
Description
![[Experiment image]](img/science/methods/toc_FX_rs.png)
Novel drugs can be rationally designed by growing and linking small molecule fragments. However, because fragments are fast and promiscuous experimentalists commonly have contradictory hits when using different techniques. In this work, we run 2.1 milliseconds of total simulation time in GPUgrid. By analyzing the trajectories with Markov state models, we are able to simultaneously predict poses, kinetics, and affinities for a library of 42 fragments against a known protease. Specifically, the target protease is factor XA, a protein involved in the coagulation pathway whose inhibition is used to treat thrombosis. The results accurately reproduced previous crystallographic, kinetic and thermodynamic data, and showed our method can be useful to recapitulate experimental data in other targets.
Publications
- N. Ferruz, M. J. Harvey, J. Mestres and G. De Fabritiis, Insights from Fragment Hit Binding Assays by Molecular Simulations, J. Chem. Inf. Model., 2015, 55, pp 2200-2205
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | Erik Postnieks | 1,037,914,550.00 |
| 2 | Stoneageman | 359,155,850.00 |
| 3 | GPUGRID | 298,344,725.00 |
| 4 | Retvari Zoltan | 298,201,600.00 |
| 5 | flashawk | 186,546,625.00 |
| 6 | Venec | 176,527,525.00 |
| 7 | Roald | 169,022,700.00 |
| 8 | Grzegorz Roman Granowski | 167,628,975.00 |
| 9 | RaymondFO* | 155,470,125.00 |
| 10 | comfortw | 124,323,425.00 |
Reranking Docking Poses Using Molecular Simulations and Approximate Free Energy Methods
WU tags: R(L|C)
Description
![[Experiment image]](img/science/methods/selections2.jpg)
Fast and accurate identification of active compounds is essential for effective use of virtual screening workflows. Here, we have compared the ligand-ranking efficiency of the linear interaction energy (LIE) method against standard docking approaches. Using a trypsin set of 1549 compounds, we performed 12,250 molecular dynamics simulations. The LIE method proved effective but did not yield results significantly better than those obtained with docking codes. The entire database of simulations is released.
Publications
- G. Lauro, N. Ferruz, S. Fulle, M. J. Harvey, P. W. Finn, and G. De Fabritiis,Reranking Docking Poses Using Molecular Simulations and Approximate Free Energy Methods, J. Chem. Inf. Model., 2014, 54 (8), pp 2185–2189
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | RaymondFO* | 14,447,300.00 |
| 2 | Erik Postnieks | 12,647,675.00 |
| 3 | Retvari Zoltan | 9,795,825.00 |
| 4 | GPUGRID | 9,135,500.00 |
| 5 | Bedrich Hajek | 9,101,550.00 |
| 6 | Venec | 8,415,500.00 |
| 7 | Thomas Huettinger | 7,125,175.00 |
| 8 | flashawk | 7,092,300.00 |
| 9 | steinrar | 6,546,661.00 |
| 10 | STE\/E | 6,091,013.00 |
On-the-Fly Learning and Sampling of Ligand Binding by High-Throughput Molecular Simulations
WU tags: BenAdapt
Description
![[Experiment image]](img/science/methods/benadapt_paper.gif)
The most important information that can be taken out of a protein-ligand binding simulation is the binding poses of the ligand, the binding pathways and the free energy of binding. However in classical sampling simulations lots of simulation time is wasted re-sampling areas of low interest which might not lie on the binding pathway or have already been sampled adequately. Therefore, we proposed a adaptive sampling method by which it is possible to sample more strongly along the binding pathway of a ligand and thus achieved a 10 times speedup on the estimation of the binding free energy of the ligand compared to classical sampling.
Publications
- S. Doerr and G. De Fabritiis, On-the-Fly Learning and Sampling of Ligand Binding by High-Throughput Molecular Simulations, J. Chem. Theory Comput., 10 (5), 2064-2069, (2014)
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | Stoneageman | 17,694,500.00 |
| 2 | RaymondFO* | 9,307,400.00 |
| 3 | GPUGRID | 8,786,400.00 |
| 4 | Retvari Zoltan | 6,772,200.00 |
| 5 | flashawk | 6,071,100.00 |
| 6 | HA-SOFT, s.r.o. | 5,639,400.00 |
| 7 | Venec | 5,520,600.00 |
| 8 | Rick A. Sponholz | 5,366,400.00 |
| 9 | dominik | 4,821,000.00 |
| 10 | Grzegorz Roman Granowski | 4,597,400.00 |
Kinetic Characterization of Fragment Binding in AmpC beta-Lactamase by High-Throughput Molecular Simulations
WU tags: 2HDQ
Description
![[Experiment image]](img/science/methods/2014_2HDQ.gif)
Small molecules used in fragment-based drug discovery form multiple, promiscuous binding complexes difficult to capture experimentally. Here, we identify such binding poses and their associated energetics and kinetics using molecular dynamics simulations on AmpC β-lactamase. Only one of the crystallographic binding poses was found to be thermodynamically favorable; however, the ligand shows several binding poses within the pocket. This study demonstrates free-binding molecular simulations in the context of fragment-to-lead development and its potential application in drug design.
Publications
- P. Bisignano*, S. Doerr*, M. J. Harvey, A. D. Favia, A. Cavalli, and G. De Fabritiis, Kinetic Characterization of Fragment Binding in AmpC beta-Lactamase by High-Throughput Molecular Simulations, J. Chem. Info. Model., 54 (2), 362-366, 2014
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | Stoneageman | 24,067,500.00 |
| 2 | GPUGRID | 14,355,000.00 |
| 3 | Retvari Zoltan | 12,825,000.00 |
| 4 | RaymondFO* | 12,420,000.00 |
| 5 | flashawk | 12,220,000.00 |
| 6 | Venec | 11,376,875.00 |
| 7 | Bedrich Hajek | 8,505,000.00 |
| 8 | Grzegorz Roman Granowski | 8,077,500.00 |
| 9 | underwater | 8,055,000.00 |
| 10 | werdwerdus | 6,401,250.00 |
Identification of slow molecular order parameters for Markov model construction.
WU tags:
Description
A goal in the kinetic characterization of a macromolecular system is the description of its slow relaxation processes via (i) identification of the structural changes involved in these processes and (ii) estimation of the rates or timescales at which these slow processes occur. Most of the approaches to this task, including Markov models, master-equation models, and kinetic network models, start by discretizing the high-dimensional state space and then characterize relaxation processes in terms of the eigenvectors and eigenvalues of a discrete transition matrix. The practical success of such an approach depends very much on the ability to finely discretize the slow order parameters. How can this task be achieved in a high-dimensional configuration space without relying on subjective guesses of the slow order parameters? In this paper, we use the variational principle of conformation dynamics to derive an optimal way of identifying the "slow subspace" of a large set of prior order parameters - either generic internal coordinates or a user-defined set of parameters. Using a variational formulation of conformational dynamics, it is shown that an existing method-the time-lagged independent component analysis-provides the optional solution to this problem. In addition, optimal indicators-order parameters indicating the progress of the slow transitions and thus may serve as reaction coordinates-are readily identified. We demonstrate that the slow subspace is well suited to construct accurate kinetic models of two sets of molecular dynamics simulations, the 6-residue fluorescent peptide MR121-GSGSW and the 30-residue intrinsically disordered peptide kinase inducible domain (KID). The identified optimal indicators reveal the structural changes associated with the slow processes of the molecular system under analysis.
Publications
- Pérez-Hernández G, Paul F, Giorgino T, De Fabritiis G, Noé F, Identification of slow molecular order parameters for Markov model construction. The Journal of chemical physics 2013. doi:10.1063/1.4811489
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | Venec | 13,468,338.00 |
| 2 | Stoneageman | 8,446,204.00 |
| 3 | mclaver | 8,205,305.00 |
| 4 | 123bob | 7,710,927.00 |
| 5 | escape | 7,447,455.00 |
| 6 | comfortw | 7,234,691.00 |
| 7 | netwraith | 7,186,950.00 |
| 8 | Helmholdt | 5,222,888.00 |
| 9 | Ketzer7 | 4,980,192.00 |
| 10 | Laboratory of Theory of Biopolymers | 4,966,325.00 |
Free binding of inhibitor benzamidine to enzyme trypsin
WU tags: TRYP, PYRT
Description
Identification of inhibitor molecules (drugs) that bind to enzymes or other proteins (targets) has been, and will be, the principal goal in drug discovery processes. Computational biologists/biochemists develop computational methods that span from ligand binding pose prediction to ligand binding affinity calculations, to aid in the quest for finding new, better and safer drugs. With our experiments, we show for the first time, a complete process of binding of a drug-like molecule to its target protein. Our molecules are used as a toy model in a proof-of-concept study for future and more relevant cases. In addition to reproduction of crystallographic ligand binding pose (also tackled by much cheaper but more coarse-grained techniques named 'computational docking'), we show the complete pathway of binding that the inhibitor follows from the solvent to the pocket where it binds. We detect several amino-acids in trypsin that consistently interact with benzamidine as it binds, which indicates that there is a prefered pathway for benzamidine to bind and therefore inhibit the function of trypsin. The principal outcome of this work is that with Molecular Dynamics simulations, it is now possible to study full binding events, being able to visualize and quantify the whole process of binding with a single computational experiment. We are confident that this achievement will allow a much deeper understanding of the processes of binding for small drug-like molecules which may then lead to the design of new, better and safer drugs.
Publications
- I. Buch, T. Giorgino and G. De Fabritiis, Complete reconstruction of an enzyme-inhibitor binding process by molecular dynamics simulations, Proc. Natl. Acad. Sci. USA 108(25), 10184-10189 (2011)
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | Stoneageman | 15,732,027.00 |
| 2 | Retvari Zoltan | 5,832,012.00 |
| 3 | escape | 5,443,161.00 |
| 4 | whizbang | 4,093,559.00 |
| 5 | tng* | 3,846,724.00 |
| 6 | Bikermatt | 3,768,642.00 |
| 7 | dataman | 3,575,314.00 |
| 8 | netwraith | 3,510,954.00 |
| 9 | comfortw | 3,481,642.00 |
| 10 | dak1640 | 3,236,762.00 |
Forward-Reverse Steered Molecular Dynamics
WU tags: GA
Description
Potassium ion permeation in Gramicidin A. We are giving workunits comprising full-atom simulations of gramidicin A for ion transport, a total of 30,000 atoms. Each workunit lasts less than one day and you have to complete it before 4 days.
Publications
- T. Giorgino and G. De Fabritiis, A high-throughput steered molecular dynamics study on the free energy profile of ion permeation through gramicidin A, J. Chem. Theory Comput.,7 , 1943–1950 (2011)
| Badge | Rank | Name | Credit |
|---|---|---|---|
| 1 | Stoneageman | 4,713,395.00 |
| 2 | dak1640 | 3,857,211.00 |
| 3 | GPUGRID Role account | 3,350,192.00 |
| 4 | Oleg Tchij | 3,212,435.00 |
| 5 | Grzegorz Roman Granowski | 3,163,300.00 |
| 6 | CNT - IQE | 2,107,085.00 |
| 7 | 123bob | 1,990,814.00 |
| 8 | X-Files 27 | 1,609,954.00 |
| 9 | kevint | 1,506,964.00 |
| 10 | netwraith | 1,353,994.00 |
