Antibody Structure Prediction Panel

Predict the variable loop (CDR) region of an antibody, including the hypervariable loops and the associated framework, by homology modeling.

To open this panel: click the Tasks button and browse to Biologics → Antibody Structure Prediction.

Using
Features
Additional Resources

Using the Antibody Structure Prediction Panel

When you have finished building an antibody structure, you might want to refine the CDR loops. You can do this in the Antibody Loop Modeling Panel.

When the results are incorporated into the project, you can use the Workflow Action Menu for the entry group in the Entry List (Entries) to choose an action to apply to the results.

Antibody Structure Prediction Panel Features

The tabs displayed depend on whether you are running a single structure prediction or multiple structures predictions in batch mode.

Step indicator
ANTIBODY FORMAT step
FRAMEWORK SEARCH step
BASIC LOOP MODELING step
BATCH LOOP MODELING step

Step indicator

The current step is highlighted in this diagram of the steps. There are three steps for single antibody modeling, two for batch modeling of multiple related antibodies.

ANTIBODY FORMAT step

Choose the antibody format and import the sequences for the regions in the format.

Build Model options

Select an option for the models to build. The choice affects the features that are displayed in the tab.

Single—build a single model. This option has the most flexibility in the choices you can make for building the model.
Batch—build models for multiple related antibodies, with input taken from CSV files.

Use ML for prediction (Beta) option

Select to utilize machine learning (ML) methods for quicker results or when standard templates are not suitable. When enabled, the Framework Search or Batch Modeling button at the bottom of the panel will change to a Generate Models button.

Note: This option supports only Fv and sdAB formats.

Format option menu

Choose the format of the antibody. The Sequences selection tools and Antibody diagram are updated according to the choice of format.

Full Monospecific— specify the sequences for the full heavy and light chains in the symmetric case where the two heavy and the two light chains are the same.
Full Bispecific—specify the sequences for the full heavy and light chains in the asymmetric case where the two heavy and the two light chains are different.
Fab--specify the sequence for the A and B regions (VL+CL, VH+CH1)
F(ab)2 Monospecific—specify the sequences for the F(ab)2 heavy and light chains in the symmetric case where the two heavy and the two light chains are the same.
F(ab)2 Bispecific—specify the sequences for the F(ab)2 heavy and light chains in the asymmetric case where the two heavy and the two light chains are different.
Fv or scFv--specify the sequence for the variable regions or single-chain variable regions only.
sdAB—single-chain antibody variable region.

Load batch file text box and Browse button

Enter the batch file name in this text box, or click Browse and navigate to the batch file. The name of the batch file you selected is displayed in the text box. The allowed file types are: CSV.

Sequences display and tools

These tools allow you to view or select the sequence regions. Selecting sequence regions can only be done when building a single model. In batch mode, you can only view the sequence regions.

Antibody diagram

This diagram shows the antibody of the chosen format. Each region has a tooltip identifying the region (VL, VH, etc.). The regions that you can specify for this format are outlined. If you chose Batch for building the model, the diagram is informational only, as the regions are set from the batch file. If you chose Single, clicking on the regions displays a menu from which you can choose the source of the sequence, from the following:

File—opens a file selector so you can locate and open a file that contains a sequence.
Workspace—use the sequence from the structure in the Workspace.
Project Table (selected)—use the sequence from the selected entry. Only one entry must be selected.
PDB ID—import a sequence from the PDB with the specified ID. Opens the Get PDB File Dialog Box.
Enter new sequence—Type or paste in the sequence as a string of single-letter codes. Opens the Sequence Editor dialog box, in which you can name the sequence and type it or paste it into the box and click Add. The name for this sequence will be displayed in the legend and in the tooltip for the selected subregion.
Recent—Import a structure from your recent file selections from under the Recent list.

The final item, Clear, allows you to clear the selection for the region.

If the structure source contains multiple chains, you are prompted to choose the desired chain.

Clear all regions link

Clear the sequences from all regions.

Load Fc region options

Load the Fc region from the specified source. These options are not present for Fv, Fab, F(ab)2, and sdAB formats. In Batch mode, these options are above the antibody diagram, and are labeled Load constant region.

From file—Load the Fc region from a structure file. Click Browse to open a file selector to locate and open the file.
From template--Load the Fc region from a template. Click Select to choose from the available templates. Only present for full antibody formats.

Load reference for domain orientation option

Load the reference for domain orientation from the specified source. This option is only present for F(ab)2 formats.

Load constant region option

Load the structure file from the specified source that contains the desired Fab constant region. This option is only present for Fab formats.

Build single-chain Fv (scFv) option

Select this option to enable the Enter linker sequence text box, where you can type in or paste a full linker sequence or monomer unit.

FRAMEWORK SEARCH step

Choose the frameworks for modeling the antibody. This step is only present if you are modeling a single antibody. In batch mode, this step is replaced by the BATCH LOOP MODELING step.

Numbering Scheme option menu

Choose the numbering scheme to use, from Chothia, Enhanced Chothia, Kabat, IMGT, AHo. The definition of the CDR loops, ranking of frameworks, CDR loop clustering, and choice of loop templates are affected by the numbering scheme. These differences can lead to different predicted structures.

Antibody database link

Select the databases that you want to use for homology modeling. These databases are the ones that are searched for templates, and are used for both the framework regions and the CDR regions of the variable domains. Opens the Choose Database Dialog Box, in which you can add databases to the list, remove them from the list, and mark them as active. All databases that you select as active are searched when you run a search for homologs.

Choose Frameworks section

Choose the framework source and load possible frameworks, then select the ones you want to use.

Framework from option menu and Run button

Choose the source of the framework coordinates. If you choose Homology search, click Run to run the homology search. The results are returned in the homologs table below. If you choose Input structure, the rest of the controls in this section are hidden.

Homology search—Use homology modeling to determine the coordinates. When this option is selected, click Run to run the homology search. The results are returned in the homologs table below.
Input structure—Use the coordinates of the input structure for the framework region. The sequences that you imported for the light and heavy regions must be associated with a structure.

Separate chains option

Select this option if you want to choose different templates for the heavy chain and the light chain. The table below is split into two tables, one for the heavy chain and one for the light chain, with the appropriate selection of table columns. You can then make separate choices for the two chains in the table. Not present if sdAB was chosen as the format.

Interface alignment structure link: Select a common template for aligning heavy and light frameworks, if you selected Separate chains and want to use a common framework other than the default (the one with the highest score). Opens the Select Common Framework dialog box, where you can choose a template from a list of available templates. Not present if sdAB was chosen as the format.

Homologs table

This table displays the results of the search for homologs, sorted by composite score. The Similarity Score is the number of matching residues divided by the total number of residues, where "matching" means that the two residues have a positive score in the BLOSUM62 matrix. Likewise, the Identity is the number of identical residues divided by the total number of residues. The tooltip displayed when hovering over a column header provides additional details on how the score is derived. Note: if the Similarity Score is a negative value, the column will display "Mismatched Proline" instead of the value.

The highest-scoring homolog is selected by default. You can select a different homolog from the table to use for the framework region.

Heavy	PDB ID of the homolog for the heavy framework region
Light	PDB ID of the homolog for the light framework region. Not present if sdAB was selected for the format.
Composite Score	Average of the Heavy Sim. and Light Sim. scores. This score is used to order the homologs in the table.
Antigen type	If the template contains an antigen, its type and chain length are listed in this column. If it does not contain an antigen, the column shows None.
Heavy Fr. Sim.	Sequence similarity of the framework region for the heavy chain.
Light Fr. Sim.	Sequence similarity of the framework region for the light chain. Not present if sdAB was selected for the format.
Heavy Fr. Iden.	Sequence identity of the framework region for the heavy chain.
Light Fr. Iden.	Sequence identity of the framework region for the light chain. Not present if sdAB was selected for the format.
Heavy Sim.	Sequence similarity of the entire variable domain sequence (framework and CDR) for the heavy chain.
Light Sim.	Sequence similarity of the entire variable domain sequence (framework and CDR) for the light chain. Not present if sdAB was selected for the format.
Heavy Iden.	Sequence identity of the entire variable domain sequence (framework and CDR) for the heavy chain.
Light Iden.	Sequence identity of the entire variable domain sequence (framework and CDR) for the light chain. Not present if sdAB was selected for the format.
PDB Resolution	Resolution of the PDB structure, if available; otherwise experimental method, if available.

Reset Selection link

Reset the selection in the table to the highest-scoring homolog.

View Alignment link

View the alignment of the chosen homolog with the input sequence in the Multiple Sequence Viewer/Editor Panel. The alignments for heavy and light chains are shown in separate tabs.

Note: The opening of the alignments in the Multiple Sequence Viewer/Editor Panel is for viewing only. If you make adjustments in that panel, they are not transferred back to the Antibody Structure Prediction workflow.

Export Results link

Export the results of the search to a CSV file. Opens a file selector, so you can choose a location and name the file.

BASIC LOOP MODELING step

Perform modeling for the CDR loops. This step is not present for batch modeling of multiple antibodies, for which the settings are made in the BATCH LOOP MODELING step.

Loop clustering progress bar

This progress bar is displayed at the top of this step on entry, and is removed when the loop clustering finishes. The antibody databases are searched for loops of the same length as those in the query sequence for each of the six loops, and the loops are clustered structurally.

Select source for CDR loops table

This table lists the CDR loops and provides buttons for selecting the source of input coordinates. There are six rows for regular predictions, and three for single-domain predictions. The table columns are as follows:

CDR Loop	Label of the CDR loop.
Loop Database	Column of radio buttons for using the antibody loop database to predict the loop.
Framework Template	Column of radio buttons for using the framework template to predict the loop. You can only do this if you chose to do a homology search for the framework.
Input Structure	Column of radio buttons to use the input coordinates for the loop. You can only do so if the antibody sequence was imported from a PDB structure with coordinates available, and if the framework region was also taken from the input structure.

Reset selection link: Reset the selection of the loop source to the default.
Details link: Show details for the loop that is selected in the table. Opens the Loop Clusters Dialog Box. This dialog box lists the clusters for each loop with information on the cluster representative (most similar structure in the cluster), and allows you to select a template to use in the model.
Include template antigen link: Include the antigen from the template when constructing the model. Antigens are listed in the template table in the FRAMEWORK SEARCH step.

Results option menu

Choose the model that you want to inspect or process.

View in Workspace button

View the model selected from the Model to view option menu in the Workspace. The model is colored by residue with the following color scheme:

Blue—residues for which the full residue conformation was copied from the template.
Cyan—residues for which the residue backbone conformation was copied from the template, and the side chain was modeled (because there was a residue mutation in the template relative to the query).
Red—residues for which both the backbone and the side chain were modeled.
Maroon—residues in the CDR loops.

Refine Loops button

Refine loops with Prime refinement. Opens the Antibody Loop Modeling Panel.

Number of models text box

Specify the number of models to generate for the antibody.

Job toolbar

Manage job submission and settings. See Job Toolbar for a description of this toolbar.

Status bar

Use the Reset button to reset the panel to its default settings and clear any data from the panel. If the panel has a Job toolbar, you can also reset the panel from the Settings button menu.

If you can submit a job from the panel, the status bar displays information about the current job settings and status for the panel. The settings include the job name, task name and task settings (if any), number of subjobs (if any) and the host name and job incorporation setting. The job status can include messages about job start, job completion and incorporation.

The status bar also contains the Help button , which opens an option menu with choices to open the help topic for the panel (Documentation), launch Maestro Assistant, or if available, choose from an option menu of Tutorials. If the panel is used by one or more tutorials, hover over the Tutorials option to display a list of tutorials. Choosing a tutorial opens the tutorial topic.

Generate Loop Models button

Run the job to generate the specified number of models of the antibody. The loops are built first, then a full side-chain sampling is done on the loops to produce the final structures.

Note

If running a job with the scFv antibody format, a final step must be performed after the job has been completed. Use the link in the job completed notification banner or choose Protein Linker Design from the Workspace Action Menu.

BATCH LOOP MODELING step

Set up the framework and choose options for batch modeling of multiple antibodies.

Apply Numbering Scheme option menu

Antibody database link

Options section

Specify options for the framework and modeling.

Framework options

Select the source of the framework region.

Choose automatically—load the framework from the structure (template) source for each antibody.
Use PDB ID—Use the specified PDB structure for the framework.

Keep glycan in Fc option

Include the glycan in the Fc region of the templates when building the model. This option is only available in Batch Mode for Full Monospecific and Full Bispecific jobs.

Include antigen option

Include the antigen from the template in the antibody structure when building each model.