An HTML 5 based diagraming tool, Gliffy is an online tool that offers 2 diagrams for free to. UML Sequence Diagrams display an interaction as a two-dimensional chart. Sequence Diagrams are used to design, document and validate the architecture, interfaces and logic of the system by describing the sequence of actions that need to be performed to complete a task. Why Creating Sequence Diagrams is Faster with Creately.
In this article, Senior Scientist Dmitry Lupyan introduces the Simulation Interactions Diagram (SID) available with Desmond in Release 2013-2. For more information, contact us. Or, look at section 5.3 of the Desmond User Manual.
Molecular Dynamics (MD) simulations are an important computational tool for understanding the physical basis of the structure and function of biological macromolecules. In contrast to a static description of protein-ligand interactions, MD simulations contain explicit treatment of water and intrinsic receptor flexibility, capturing the dynamic nature of protein-ligand interactions. These advantages make MD a powerful tool for investigating atomic-level interactions, which are at the heart of ligand affinity and selectivity towards its target. The information extracted from such analyses can be used to better predict binding models and rationalize structure-activity relationship (SAR) data.
Available with Schrödinger Release 2013-2, we are pleased to announce the Simulation Interactions Diagram or SID, a new post-MD analysis tool for exploring protein-ligand interactions. SID integrates molecular analysis utilities with novel plotting and visualization tools to obtain unparalleled insights into atomic-level protein-ligand interactions.
Paradigm Shift in MD Analysis
Analyzing typical MD simulation data requires substantial technical skills in addition to intimate knowledge of the molecular system being studied. This knowledge, in combination with viewing and animating the trajectory, can be used to create a hypothesis of how a drug-like molecule interacts with a protein. However, movements at an atomic-level are so rapid, complex, and occurring on different timescales, that it is very difficult to identify – and easy to overlook – important events underlying the interactions. In addition, in the new age of dynamics-aided virtual screening (MDVS), manually analyzing multiple systems of several different compounds can quickly become an overwhelming and time-consuming task. SID is used to automate analyses after a MD simulation is complete. These results are then organized in the SID panel, with plots and diagrams for easy analysis.
In addition to providing interactive plots and diagrams for exploring the data, the SID panel can also be used to output:
- A PDF report: A multi-page PDF report, containing all the results within the panel and a full description of the molecular system. All the plots and diagrams are accompanied by detailed explanations. These reports can easily be shared with colleagues or stored in electronic notebooks for further reference. An example report from an NPT simulation with FABP4 protein in complex with ibuprofen (PDB code: 3p6h) can be viewed here.
- Images of all plots: All the visual elements and plots are easily exported into PNG and SVG image formats, providing a means to quickly insert these images into a manuscript or a slideshow presentation.
- Raw data: All the results can be exported into a text file for further processing and plotting with third-party tools.
Exploring Post-MD Protein-Ligand Systems with SID
To analyze protein-ligand MD trajectories, Schrödinger scientists decided to incorporate several useful types of analyses into SID. The initial release of SID includes the analytical tools detailed below, and there are immediate plans to include several new tools in upcoming releases. For the figures shown on the next few pages, results from analyses performed on a 10-ns simulation of FABP4 protein in complex with ibuprofen (PDB code: 3p6h) are illustrated.
Protein and Ligand RMSD (PL-RMSD)
Root mean square deviation (RMSD) of the protein and ligand, with predefined atom selections are pre-computed and displayed in the PL-RMSD window tab (Figure 1). Monitoring RMSD of the protein can give insights into its structural conformation throughout the simulation, providing an indication of the stability of the protein and whether the simulation has equilibrated. Ligand RMSD can indicate how stable the ligand is with respect to the protein, as well as the evolution of its internal conformation.
Figure 1: The PL-RMSD Tab shows Protein and Ligand RMSD values (left and right y-axes, respectively) plotted against simulation time. The RMSD results of different atom selections (C-alphas, backbone, side chains, heavy atoms) can be viewed separately using the toggles at the top of the panel.
Protein RMSF (P-RMSF)
Root mean square fluctuations (RMSF) of protein residues are displayed in the P-RMSF window tab (Figure 2), enabling visualization of segments along the protein that fluctuate the most during the simulation. Typically these fluctuations should correlate with the experimental x-ray B-factor, which can be toggled on and off for easy comparison with experimental data.
In order to explore which protein residues come into contact with the ligand, SID allows you to highlight these residues. Additionally, since secondary-structure elements are more rigid than the unstructured loop region(s), this panel allows you to see how fluctuations correspond to secondary structure elements in your simulation by overlaying alpha-helix and beta-strand regions of the protein onto the plot.
Figure 2: The Protein RMSF tab shows protein residue fluctuations from the entire simulation (light blue curve); experimental B-factor is shown as a red curve; residues that are in contact with the ligand are indicated by green vertical lines; salmon and cyan rectangles show alpha-helix and beta-strand regions, respectively; the blue vertical line indicates the currently selected residue.
Ligand RMSF (L-RMSF)
Ligand root mean square fluctuations (RMSF) are displayed in the L-RMSF window tab (Figure 3). A ligand’s atom fluctuations can provide insight into how ligand fragments interact with the protein, and into the entropic role of the ligand in the binding pocket.
Figure 3: The Ligand RMSF tab shows a ligand‘s atom fluctuations throughout the simulation. Hovering over the plot highlights the currently selected atom with a blue vertical line, displays relevant information for the atom, and highlights the atom in the 2-D schematic.
Protein ligand interactions can be explored in either the PL-Contacts or LP-Contacts window tabs (Figures 4 and 5). Various specific and nonspecific protein-ligand interactions are monitored throughout the simulations and presented in these protein- and ligand-centric tabs. Here, interactions are categorized into four types: hydrogen bonds, hydrophobic, ionic, and water bridges. Using the SID panel, these interaction types can be further broken down into several subtypes.
User Interaction Diagram
In Figure 4, the protein-centric results (PL-Contacts) are shown in a histogram. All residues that come into contact with the ligand throughout the trajectory are shown here and are color-coded by interaction types. In addition, the SID panel displays a timeline representation of these interactions where different types of interactions can be turned on and off to explore their occurrence in the simulation.
Figure 4: The Protein-Ligand Contacts tab summarizes a number of protein-centric interactions. On the top, a histogram displays the type of protein-ligand interactions; on the bottom, a plot provides a timeline representation of the same contacts, which can help with visualizing when these interactions take place.
Concrete Interaction Diagram
In Figure 5, the ligand-centric tab (LP-Contacts) provides a summary of all the contacts in the form of an interactive 2-D diagram. Interaction strength is quantified by the frequency of occurrences in the trajectory, making it easy to filter out rarely occurring events and focus on more important interactions.
Figure 5: The Ligand-Protein Contacts tab summarizes the interactions that occur during the course of the simulation. Different interaction types are indicated by color and line type. A slider can be used to decrease or increase the cutoff value for displaying interaction occurrences.
To gain insights into ligand conformational evolution over the course of a trajectory, the L-Torsions plot monitors all ligand rotatable bonds (Figure 6). In the panel, a 2-D diagram of the ligand is shown, with color-coded rotatable bonds. Each dial plot summarizes the conformational changes of a rotatable bond as a function of time, and each distribution plot summarizes the bond’s conformational distribution.
Figure 6: The L-Torsions tab allows for analysis of ligand rotatable bonds. Each rotatable bond is color-coded, as shown on the 2-D ligand structure. Each dial plot illustrates the conformational changes of a bond over time, and each distribution plot includes a histogram illustrating the distribution of the bonds’ conformations. A force field potential curve corresponding to the torsions of the given rotatable bond is also included in each distribution plot.
Object Interaction Diagram
SID is bundled with Desmond in Schrödinger Release 2013-2. If you are interested in learning more or if you would like to request a demo, please contact your Account Manager or fill out our Trial Request form. We are continuing our product development efforts and look forward to receiving your feedback to guide future enhancements to SID and related MD analysis tools.