Peptide Cyclization

Tutorial Created with Software Release: 2026-2
Topics: Biologics Drug Discovery, Peptide Design
Products Used: BioLuminate, Prime

Tutorial files

18.9 MB
This tutorial is written for use with a 3-button mouse with a scroll wheel.
Words found in the Glossary of Terms are shown like this: Workspacethe 3D display area in the center of the main window, where molecular structures are displayed

 

Tip: You can hover over a glossary term to display its definition. You can click on an image to expand it in the page.
Abstract:

This tutorial demonstrates how to cyclize peptides using the macrocylize.py script via the command line interface. It also covers rank-ordering the cyclized peptides using the Prime MM-GBSA method.

Tutorial Content
  1. Introduction to Cyclization of Peptides
  2. Creating Projects and Importing Structures
  3. Analyzing Protein-Peptide Interactions
  4. Cyclizing Linear Peptides
  5. Rank-ordering the Cyclized Peptides Using Prime MM-GBSA
  6. Conclusion and References
  7. Glossary of Terms

1. Introduction to Cyclization of Peptides

Cyclization is a widely used strategy to optimize properties of linear peptides by converting a flexible chain into a conformationally restricted macrocycle. Linear peptides can adopt many conformations in solution, and this flexibility often reduces binding affinity. Cyclization limits this flexibility, stabilizing the bioactive conformation, and frequently improving properties like potency and selectivity. Cyclization can also contribute to formation of intramolecular hydrogen bonds. These hydrogen bonds shield the polar groups from the solvent, which results in decreased polarity of the peptide, ultimately enhancing membrane permeability. In addition, cyclization also increases resistance to enzymatic degradation by making cleavage sites less accessible to proteases.

Maestro offers several methods to cyclize linear peptides. These methods are briefly described below:

  1. Sidechain Bridge Cyclization – This method uses macrocyclize.py script, which automates the exploration of macrocyclic peptide design space by systematically generating and evaluating cyclizations of peptides. This script has two modes: small molecule (-mode sm) and peptide (-mode peptide). The small molecule mode is used to cyclize/bridge between two defined attachment points. The peptide mode performs a similar reaction but it enumerates over all peptide residues using pre-defined chemistries (disulfide bonds, lactam bridges, thioether linkages, etc.) in an existing peptide, mutates them to a residue compatible for cyclization and tries to perform the cyclization reaction. Similar cyclizations can also be done in the small molecule mode, offering finer-grained control of the reaction. However, the peptide mode is optimized for common peptide cyclizations. This method is particularly useful for designing stapled peptides, lactam-linked peptides, or disulfide cyclized variants. The automated and high-throughput nature of this method makes it well suited for screening many residue pairs and linker options efficiently, helping prioritize promising candidates for experimental evaluation. Broadly, the protocol automates the following steps:

  2. Ligand Designer Macrocyclization – This method provides an interactive, structure-based approach for cyclizing small molecules and peptides directly within the context of a protein binding site using the Ligand Designer panel. Starting from a co-crystal structure or a reliable docked pose, the bound ligand is analyzed to identify solvent-exposed regions where a linker can be introduced to form a macrocycle while preserving key receptor interactions. The graphical interface allows you to visually inspect candidate attachment points, generate macrocycles with different linker chemistries, and immediately evaluate how each design fits within the binding pocket. Because the receptor environment is considered throughout the process, this method is especially useful for maintaining critical contacts and avoiding steric clashes. It is well suited for projects where one wants to rapidly explore multiple cyclization strategies, optimize the linker of an existing macrocycle, or convert a linear peptide into a more stable cyclic analog.
  3. Manual Cyclization – Manual cyclization using the 3D Builder offers the greatest flexibility for creating custom cyclic peptides. In this method, you can select any two atoms in the peptide and form a bond between them, while Maestro handles the valence adjustment. This method is useful for constructing unusual linkages not available in predefined bridge libraries. Additionally, if your peptide has less than 200 atoms, then you can also load it in the 2D sketcher to perform manual cyclization. Note that since the bond is manually created, the resulting structure should be followed by energy minimization to remove steric clashes and correct geometric distortions. Further, the cyclized structure should be carefully inspected since this method does not perform automated enumeration, strain evaluation, or chemistry validation.

This tutorial uses the 5VAI structure, which has Glucagon-like peptide 1 (GLP-1) bound to the GLP-1 receptor (GLP-1R) in complex with the Gs protein. In this tutorial, you will learn how to cyclize peptides using the macrocyclize.py script via the command line interface. You will also learn how to rank-order the cyclized peptide structures using the Prime MM-GBSA method.

2. Creating Projects and Importing Structures

  1. Open BioLuminate and create a new project named cyclize_peptides.prj for this tutorial.
  2. Download the tutorial zip file including input files and reference outputs here: Tutorial Files
  3. After downloading the zip file, unzip the contents in your Working Directorythe location where files are saved for ease of access throughout the tutorial.

Figure 2-1. Importing the prepared complex.

To save time, the 5VAI structure has been prepared using the Protein Preparation Workflow and included in the tutorial files.

  1. Go to Files > Import Structures.
  2. Navigate to the 5VAI_prepared-out.maegz in your Working Directorythe location where files are saved.
  3. Click Open.
    • The prepared structure is loaded into the Workspacethe 3D display area in the center of the main window, where molecular structures are displayed.

3. Analyzing Protein-Peptide Interactions

Protein-peptide interactions are critical for many biological processes like molecular recognition and signal transduction. Typically, short linear peptides interact with protein receptors through hydrogen bonds, salt bridges, and pi-pi interactions. Before performing cyclization, it is important to identify the peptide residues interacting with the protein, as it is highly likely that these residues could be responsible for receptor recognition and activation.

The following interactive tutorials demonstrate how to identify peptide-protein interactions using the Protein Interaction Analysis panel and visualize these interactions in the 3D Workspace.

The N-terminal of the GLP-1 peptide has several key residues that interact with the GLP-1R receptor through hydrogen bonds, salt bridges, and pi-interactions, including GLU 9, THR 11, THR 13, ASP 15, SER 14, SER 18, etc. This publication reports that initially, the C-terminal of GLP-1 binds to the extracellular N-terminal domain of the GLP-1R. Once the C-terminal of GLP-1 is anchored, the N-terminal of GLP-1 goes into the receptor’s core that switches the receptor into its active state. These findings allow us to make informed selection of C-terminal GLP-1 residues for introducing cyclization.

Note

While the example used in this tutorial benefitted from the available 3D structure of the protein-peptide complex, this may not be the case for many practical modeling projects. When an experimental or predicted 3D structure of the peptide is not available, direct analysis of binding interactions is not possible, which can make selection of cyclization sites more challenging. In such cases, one can rely on available sequence-based information. For example, known binding peptides or previously tested mutants can be compared with non-binding sequences to identify conserved residues or short sequence motifs that are likely important for target recognition and activation. Even without structural data, these recurring patterns often indicate positions that contribute strongly to binding and therefore should be avoided when introducing cyclization. Residues outside these critical regions are generally better candidates for cyclization. Further, if additional structural insight is needed, protein-peptide docking can be used to to predict possible binding poses and identify interaction hotspots.

4. Cyclizing Linear Peptides

Many receptor activating peptides adopt helices, turns, or extended conformations upon binding. The primary goal of cyclization is to introduce conformational constraint while preserving the interactions required for binding and biological activity. In other words, cyclization is most successful when it stabilizes the bioactive conformation rather than forcing an alternative fold. This is especially important for peptides like GLP-1 which function as a receptor agonist, because the bound pose must not only retain affinity but also stabilize the receptor in its active signaling state. Here are a few key points to consider while cyclizing a peptide:

  • Residues forming key interactions with the target should generally be avoided as cyclization anchor points unless the linker can be introduced without perturbing their geometry.
  • Residues pointing away from the receptor surface, solvent-exposed, or flexible terminal segments are often better candidates for cyclization.
  • The selected residues should be spatially close enough in the bound conformation to be bridged by a chemically realistic linker without introducing excessive strain.
  • Both the linker length and attachment vector matter – a linker that is too short can distort the peptide backbone, while the one that is too long may fail to rigidify the peptide.

What if I don’t have a 3D structure of the peptide?

In this tutorial, we are using the 5VAI structure, which already contains the GLP-1 linear peptide bound to the target receptor (GLP-1R). If you do not have a 3D structure of the peptide of interest, you can generate the 3D structure using either of these options before proceeding with the cyclization workflow (only applicable for modeling peptides with natural amino acids):

Please note that regardless of the method used for modeling the 3D peptide structure, the outputs are geometric constructs and not chemically realistic structures. To ensure correct protonation states of standard ionizable side chains/termini, and eliminate any strain, clashes, or unfavorable geometries, we recommend preparing the structure using the LigPrep panel before proceeding with the cyclization workflow. To learn how to prepare cyclic peptides using the LigPrep panel, please consult Preparing Cyclic Peptides and Aligning them to a Reference Structure tutorial. 

This section uses command line tools from the Schrödinger suite. For operating system-specific instructions on getting started using the command line interface, see Running Schrödinger Applications from the Command Line.

For peptide cyclization, you generally need the following:

  • -inp_mae: Target complex which contains the target receptor and the (linear) peptide. Please note that the peptide does not have to be necessarily linear as this script can be used to introduce additional cyclizations into an existing cyclic peptide.
  • -target_asl: Target atoms for cyclization. In the case of peptides, residues work very well. The default ASL is ‘ligand’ so that the entire molecule is considered.
  • -mode: Set to peptide with -mode peptide.
  • -out_fmt: Used to configure the output file format; e.g. PV, ligand only, etc.
  • -sidechain_bridges: Use numbers to specify built-in cyclization reactions; numbers are identified with the -show_bridges_and_spacers argument.
  • -sidechain_bridge_smiles: Provide custom SMILES strings to act as bridges for cyclization.
  • -sidechain_bridge_attachments: Specify custom attachment stereochemistry during cyclization i.e. what residue type (L/D) will a bridging residue be mutated to.

Note that if neither -sidechain_bridges nor -sidechain_bridge_smiles are specified, all builtin sidechain bridges are used by default.

Additional parameters for cyclization can be fine tuned with a parameter file or by specifying individual overrides through the command line interface. To view all available parameters for both modes, you can run:

$SCHRODINGER/run -FROM psp macrocyclize.py -write_param_file

This returns:

creep_delta_dist: 3.5

creep_max_mini_steps: 50

creep_min_dist: 4.0

creep_strength: 50.0

dont_break_interactions: true

dont_break_rings: true

limb_dist_tol: 0.4

limb_distance_index: 1

max_block_ratio: 1.2

max_clash: 3

max_dist_ratio: 1.2

max_enumerate: 500000

min_block_ratio: 0.5

min_bond_dist_ratio: 2.0

min_dist_ratio: 0.5

min_num_bonds: 7

min_orient_agreement: 0.0

n_linker_confs: 50

strain_limit: 150.0

Individual cyclization parameters can be adjusted with -parameter. For more comprehensive changes, -param_file can be used to provide the full YAML parameters configuration file. Useful parameters to tune include:

  • strain_limit – filter for removing strained linkers after the cyclization reaction; in kcal/mol
  • creep_max_mini_steps – number of minimizer steps to attempt closing a cyclization
  • creep_delta_dist – the target distance to move during each step of the minimizer cycle closure

The following interactive tutorial guides you how to cyclize peptides using the macrocyclize.py script (command help). If you do not wish to run the job yourself, the pre-generated results are included in the tutorial files. To view the pre-generated results, go to File > Import Structures and choose 5VAI-prepared-out.cyclizations_ligsonly.maegz.

The cyclization yielded 86 cyclized peptides. Please note that these structures are not ranked in any particular order.

5. Rank-ordering the Cyclized Peptides Using Prime MM-GBSA

The macrocyclize.py script can generate numerous feasible cyclizations but output structures are not ranked in any particular order. There are several tools in the suite to filter compounds prior to more expensive affinity predictions by FEP+ calculations:

  • RMSD calculation by the Macrocycle Bioactive Conformation Propensity panel – The propensity for a cyclized peptide to adopt the same conformation as the linear peptide can be determined by the RMSD of a set of atoms that are considered critical for the activity. The Macrocycle Bioactive Conformation Propensity panel also allows you to generate conformations using Prime Macrocycle Conformational Sampling (Prime-MCS) as part of the job. Please note that the accuracy of the prediction is highly dependent on the macrocycle’s size, as Prime-MCS may not accurately sample conformations for large peptides.
  • Conformational stability by Molecular Dynamics Simulations – MD-based workflows like replica exchange (REST) or metadynamics simulations allow you to sample the conformational landscape of the cyclized compounds and compare it to the linear peptide.
  • Enthalpic scoring with MM-GBSA – The molecular mechanics with generalized Born and surface area (MM-GBSA) method can be used to approximate the free energy of binding between a protein and the ligand (in our case, the peptide). For detailed information on the MM-GBSA implementation in the Schrodinger suite and the calculated energy terms, please consult this knowledge base article and Re-scoring Docked Ligands with MM-GBSA tutorial.
  • Enthalpic scoring with Glide – You can use Glide to score the cyclized ligands in-place, omitting the conformational sampling, and use the resulting scores (e.g. Glide energies) as a filter for prioritizing compounds for FEP+ calculations.

The following interactive tutorial guides you to load the results of cyclization and set up the MM-GBSA calculation to rank-order the cyclized peptides based on MMGBSA dG Bind.

6. Conclusion and References

In this tutorial, you learned how to cyclize peptides using the macrocyclize.py script via the command line interface. You also learned how to rank-order the cyclized peptide structures using the Prime MM-GBSA method.

6. Glossary of Terms

Entries - a simplified view of the Project Table that allows you to perform basic operations such as selection and inclusion

Included - the entry is represented in the Workspace, the circle in the In column is blue

Project Table - displays the contents of a project and is also an interface for performing operations on selected entries, viewing properties, and organizing structures and data

Recent actions - This is a list of your recent actions, which you can use to reopen a panel, displayed below the Browse row. (Right-click to delete.)

Scratch Project - a temporary project in which work is not saved, closing a scratch project removes all current work and begins a new scratch project

Selected - (1) the atoms are chosen in the Workspace. These atoms are referred to as "the selection" or "the atom selection". Workspace operations are performed on the selected atoms. (2) The entry is chosen in the Entries table (and Project Table) and the row for the entry is highlighted. Project operations are performed on all selected entries

Working Directory - the location where files are saved

Workspace - the 3D display area in the center of the main window, where molecular structures are displayed