QuValent: Covalent Ligand Design¶
Overview¶
Unlike non-covalent ligand binding, covalent ligand binding is generally considered a two-step process. In step one of the process, the ligand adopts a favorable non-covalent orientation in the binding site. In step two, a covalent linkage is formed between the pre-oriented ligand and a residue of the binding site.
The energetics of the non-covalent step are analogous to those for any non-covalent binding event. The energetics of the second step depend upon whether the binding is irreversible or reversible. For irreversible binding, the covalent bond forms and is effectively permanent. For reversible binding, the covalent bonding event can subsequently be reversed and the ligand can dissociate from the receptor. Computational characterization of these different types of covalent inhibitors is carried out using different tools.
In the case of the irreversible binder, one typically calculates both the relative standard Ki of binding of the unreacted ligands, followed by transition state analysis of the covalent bond formation, from which the relative forward activation energy (Kinact) can be ascertained. Most binding studies of irreversible inhibitors determine an effective KI, which is Ki/Kinact.
In the case of the reversible binder, one has two options: either calculate the overall effective binding by calculating both the relative Ki for the orientation process and then the relative Ki for the reversible covalent binding; or else calculate just the net Ki for the net process of binding. Unless there is a specific reason to focus on the formal two-step approach, in most cases, the latter process suffices.
Calculating the transition state energy, important for ligands that bind irreversibly, is entirely the province of quantum mechanics. This energy cannot be reliably determined using classical mechanics. Calculating the relative free energies of binding of two different covalently bound ligands can, in principle, be determined using classical mechanics. In practice, however, this relies on the parametrization of non-canonical amino acids (commonly components of the reactive cysteine residue bound to the ligand), which often yields poor results. Quantum mechanics, on the other hand, does not rely on system-specific parametrization and can be readily applied to these non-canonical residues.
The QuValent platform enables three types of calculations:
Relative free energies for non-covalent ligand/receptor binding
Relative free energies for covalent ligand/receptor binding
Automated transition state identification and energetic analysis of the reactant-TS-product process
Depending on the particulars of the covalent ligands of interest, including their binding mode, the user may use some or all of these tools.
Calculation Choices¶
When creating a new Batch calculation, you can choose either a QuValent Reaction, QuValent FEP, or standard QUELO Free Energy Perturbation (FEP), as shown above.
QuValent Reaction: Determine the transition state energetics associated with covalent bond formation. Applicable to both reversible and irreversible ligands (though it is more commonly applied for the latter).
QuValent FEP: Determine the relative free energy of binding for a series of covalently bound ligands. Primarily applied to covalent ligands that bind reversibly, but can sometimes yield predictive correlation for a series of irrersible covalent ligands with similar warheads.
Free Energy Perturbation: Run standard non-covalent FEP.
QuValent: File Uploads¶
A covalent reaction is defined by providing a structure of the covalent ligand bound to the receptor. From that input complex, the platform will automatically determine the transition state (TS) and the energy barriers K3 (Kinact) and K4. The ligand/receptor complex helps define the binding geometry. The user can optionally provide additional ligands (in SDF format). In this case, a TS calculation will be performed for each of the input ligands. For each additional ligand, superposition upon the reference ligand will be performed to provide a starting point for the ligand/protein complex, and then K3/K4 will be determined for that ligand, as well.
File Uploads¶
Covalently Modified Protein: The receptor protein that presents the nucleophile to which the ligands will bond covalently. This must be in PDB format. The structure must contain a covalently bound ligand.
Reference ligand: A ligand residue that demonstrates non-covalent binding to the target protein. This is the ligand before the covalent bond formation occurs (and before any leaving group is removed). This structure is used to orient any additional ligands.
All ligands: Any additional ligands to be considered. Maximum common substructure (MCS) superposition will be used to find the best superposition to the reference ligand for each ligand, before creating a covalent bond and carrying out the TS calculation.
Reactions (EXPERT MODE): This upload option only appears if you have expert mode enabled. It allows you to upload one or more JSON format files that define covalent reaction types that are not built into the default platform. If such reaction definitions are required for the chemistry of the ligands you are interested in, QSimulate will work with you to define them. Once you have a reaction definition file, it can be uploaded any time you use the panel and need to use it. Note that uploaded reaction definitions are only associated with the batch calculation you are currently setting up.
Once you upload the files, this will be indicated in the fields below each Browse/Upload dialog. The SMILES representation of each ligand in the input ligand file will appear on a separate line. If you wish to remove ligand(s) from an upload, click on the red “x” at the right side of the line(s) corresponding to that/those ligand(s). You can also delete the uploaded protein PDB file (if you wish to change it) by using the red “x” to the right of that field. If a problem occurs while importing any file, an informative string will appear in the “information” field.
Ligand-Nucleophile Definition¶
In addition to uploading the receptor protein structure with a covalently bound ligand in the binding site, you must define the residues that take part in the covalent bond. This includes the ligand residue code, the nucleophile residue code, and the code for the base catalyst (if any).
Ligand Code: This is the definition of the ligand residue in the ligand/protein complex. The format is Chain_Name:Residue_Name:Residue_Number
Residue Code: This is the definition of the (nucleophile) residue of the receptor to which the ligand above is covalently attached. The format is Chain_Name:Residue_Name:Residue_Number.
Base Catalyst Residue Code: This is the residue in the receptor that is catalyzing the base catalysis. Format is Chain_Name:Residue_Name:Residue_Number. Currently, only HIS (HID/HIE/HIP) residues can be specified here. If there is no base catalyst, leave this field blank.
Initialize¶
Once you have uploaded the structural information and provided the additional values required to define the ligand, protein nucleophile, and base catalyst (if any), pressing the Initialize button will start pre-processing to create initial models for the covalently bound products for all imported ligands. This is accomplished by performing structure alignment of each of the ligands with the reference ligand and associating the resulting complex with one or more of the available reaction types (see Reaction Selection below).
When this part of the setup is finished, the “Reaction Selection” and “Ligand Selection” parts of the panel (below) will populate.
Reaction and Ligand Selection¶
Reaction Selection¶
Once Initialization is complete, the Reaction Selection and Ligand Selection portions of the panel will populate. The Reaction Selection portion of the panel presents a list of the covalent reaction definitions known to the platform. A variety of reactions are pre-populated for every calculation. If the user has specified any additional reactions by uploading json-file reaction definition files supplied by QSimulate, then these will also appear in the list.
During the Initialization process (above), each ligand was assessed for compatibility with the list of known reactions. Depending on the geometry and composition of both the ligand/nucleophile complex and the surrounding residues, only a subset (and sometimes none) of the available reactions will be applicable to a particular ligand.
Beside each reaction, a match status bar and fraction are presented, which reflect the number of ligands for which this reaction is applicable. Clicking on any reaction will populate the panel to the right with a 2D representation of the reaction process, with the electron flow indicated.
Ligand Selection¶
Clicking on any reaction in the Reaction Selection table will result in annotation for all ligands in the Ligand Selection part of the panel, indicating whether the reaction applies to that ligand. The Ligand Selection table has one line for each of the uploaded ligands. The Message column indicates whether the chosen reaction can apply to that ligand. If the chosen reaction cannot apply, the message, “No valid alignments found” appears in the message column. If the reaction can apply, the Message column is blank for that ligand.
To the left of the name of each ligand you will find a annotation: A red “x” means the ligand is not compatible with the chosen reaction. A horizontal line means you have chosen to skip this ligand (no calculation will be peformed for this ligand, even if it is compatible with the reaction). A green check mark means that this ligand is selected for the calculation.
For example, in the screenshot above, the selected reaction applies only to ligands “RM4” and “RM7”. A green checkmark appears to the left of each of these, which indicates that a transition state calculation will be preformed for both. The reaction chosen does not apply to either the REF or RM3 ligands, and so red x marks appear to the left of those ligands. Selecting a different reaction will result in a different set of annotation.
You may choose to select as many of the compatible ligands as desired. A transition state calculation will be performed, separately, for each selected ligand.
If you click on the row of any ligand, that is compatible with the chosen reaction, it will be selected for inclusion (a green check appears) and a pop-up appears:
In this popup, on the left you will see the covalently bound ligand model in the context of the protein. The ligand and nucleophile are shown as ball-and-stick models. The protein is shown in a faded grey representation. On the right side of this popup you will find a list of structures. This enumerates all possible reactive warheads on the ligand. Most ligands will only have one possible reactive warhead, and so only one structure will be in the list (and it is automatically selected). If multiple possible warheads are found for the ligand, then each appears as a “structure” entry in the table on the right. If there are multiple possibilities, the user can click on the lines in this table to change the contents of the viewer, to see where the warheads are located on the molecule. The user can choose one or more of the checkboxes on the right hand side to request that transition state calculations incorporating the corresponding warheads are performed.
Options (EXPERT MODE)¶
In expert mode, the Options section of the panel is visible. This allows the user to set the functional and basis set used for the DFT calculations that performed during optimization of the transition state, reactant, and product.
Functional: The DFT functional. Functional options are described in detail in the Quantum Mechanics Background section of the manual. Ignored for portions of the calculation that are performed at the semiempirical (GFN-xTB) level.
Basis Set: The basis set to be used for DFT calculations. Basis set options are described in detail in the Quantum Mechanics Background section of the manual. Ignored for portions of the calculation that are performed at the semiempirical (GFN-xTB) level.
Reset to Default: Resets all options to their default values. If you have changed any options from their defaults, A pop-up box will appear to confirm you wish to proceed.
QuValent: Start Simulation¶
Once you are satisfied with your quantum region selections, you can start the simulation using the Start Simulation button at the bottom right of this part of the panel.
QuValent: Simulation Status¶
This portion of the panel provides the status of the calculation once it has been started using the Start Simulation button. It also provides the ability to stop and restart a calculation that has been submitted.
Stop: Stop a calculation that was previously submitted and is in progress. A stopped calculation is saved in the cloud storage associated with your account, and can be restarted later, using the “Resume” command.
Resume: Run a previously Stopped job.
Below, and also to the right of the control buttons, you will find information about the status of your job. The total estimated virtual CPU usage (vCPU) is provided, which is updated automatically every 3 minutes or can be manually updated with a press of the “here” button.
Status for each portion of the calculation is provided by bar graphs to the right of this section.
Path Generation: Evaluation of the initial path guess, using the Generalized String Method.
Refinement: Refinement, at the DFT level of theory, of the Transition State guess, and of the reactant and product
Analysis: Analysis of results of the refinement
A Summary at the bottom of this section will list the total number of molecule calculations that have completed.
Results¶
Once the calculation is complete, the results will appear in the bottom portion of the panel. Here, the energetics of the reactant, TS, and product are presented in tabular form. In addition, the atomic structures corresponding to the reactant, transition state, and product can be viewed on the right hand side of this section.
A list of ligands appears, one per row, with the name of the ligand appearing for each ligand. Clicking on any row will open the results for that ligand. Results for each ligand appear separately in a table and viewer for that ligand.
Results Table: In this table, the energies of the reactant, transition state and product are shown, in kcal/mol. The reactant is arbitrarily assigned the value of 0.00.
3D Structure Viewer The Structure viewer appears on the right side of the results portion of the table, and displays structure of the entities making the covalent bond (ligand+nucleophile) for the node selected in the reaction pathway. The structure can be rotated by using the mouse while holding the [left-click] button. The structure can be zoomed by rotating the center wheel of the mouse.
Below the 3D viewer are three buttons, for Reactant, TS, and Product. Each button changes the contents of the display to reflect the structure described by the button.
Export Results: This button can be used to export results for all the ligands to a .csv format file. The (reactant,transition state and product) structures will also be downloaded in the resulting .zip file.