Welcome to the Utrecht Biomolecular Interaction Web Portal >>

Manual

The Prodigy manuals site includes the following sections:

PRODIGY (PROtein binDIng enerGY prediction) is a web application to predict the binding affinity of protein-protein complexes based on intermolecular contacts.

Input

The protein-protein complex

PRODIGY server takes as input the three-dimensional structure of the protein-protein complex in PDB or mmCIF format:

You can choose one of the following options:

-upload the 3D structure coordinates in PDB or mmCIF format
-provide a protein databank ID code for automatic retrieval from the Protein Data Bank
- upload an archive file (.tar, .tgz, .zip, .bz2 or .tar.gz) for analyzing multiple structures at the same time
(i.e. models derived from docking simulations)

Specify the chains

The user is required to specify the chain identifiers for the molecules involved in the interaction.
Interacting molecules associated with multiple chain IDs should be provided as comma-separated list.
For the FAB/HIV-1 capsid protein p24 complex (PDB code 1E6J) the input would be expected in following format:

Interactor 1: P
Interactor 2: L,H

In this example, the binding affinity will be predicted for the interface made between chain P (p24) and chains L and H (FAB).

Please note! PRODIGY only supports the 20 standard amino acids

Archive file

When submitting an archive file to PRODIGY, make sure that the chain IDs of ALL structures are compatible with the Interactors specified in the submission form.

Parameters (optional)

Temperature (in ℃)

By default the value of the dissociation constant (Kd) is calculated at 25 ℃. The user can change this value to any desired temperature.

Job ID

The user can specify a personalized Job ID to identify the run.

Email

If an email is provided, a link with the results will be sent when the job is done. The results will be stored for 2 weeks.

Output

PRODIGY outputs are displayed online and remain downloadable for 14 days. A link to the online resource is also emailed to the user, if an email is provided.

The results returned by the server include:

1. Predicted value of the binding affinity (ΔG) expressed in kcal mol-1
2. Calculated value of the dissociation constant (Kd) at a given temperature (25 ℃ by default), expressed in Molar (M)
3. Number of intermolecular contacts (ICs) at the interface within the threshold distance of 5.5 Å, separately listed according to the contact property
4.Percentage of the charged and apolar non-interacting surface (NIS%) of the complex
5. Downloadable table (.txt) listing the residues in contact within the given threshold
6. Downloadable ready-to-run Pymol (Delano Scientific, 2002) script (.pml) with different color coding for the interacting residues
7. Archive file including all the output files

How to run PRODIGY locally

In case you have many structures on which you wish to run PRODIGY, we distribute the standalone version in form of a Github repository.
The link to the repository can be found at the BonvinLab software page.

PRODIGY(PROtein binDIng enerGY prediction - LIGands) is an automatic web server for the fast prediction of binding affinity of protein-small ligand (such as drugs or metabolites) complexes. Being a structural-based approach, it is suitable for every kind of protein-small ligand complex .

Input

The protein-small ligand complex.

PRODIGY server takes as input the three-dimensional structure of the protein-ligand complex in PDB or mmCIF format:

You can choose one of the following options:

-upload the 3D structure coordinates in PDB or mmCIF format
-provide a protein databank ID code for automatic retrieval from the Protein Data Bank
- upload an archive file (.tar, .tgz, .zip, .bz2 or .tar.gz) for analyzing multiple structures at the same time (i.e. models derived from docking simulations)

Specify the chains

The user is required to specify the chain identifiers for the molecules involved in the interaction, both for the Protein and the Ligand as well as the Residue identifier for the ligand. Example: in the complex of MDM2 with a pyrrolidine MDM2 inhibitor (PDB code 4JRG) the input would be expected in the following format:

Protein Chain ID: A
Ligand ID: A:I09

In this example, the binding affinity will be predicted for the interaction between the protein and the ligand present in chain A.

Please note! PRODIGY-LIG only supports the 20 standard amino acids for the Protein Chain

When submitting an archive file to PRODIGY, make sure that the chain IDs of ALL structures are compatible with the Interactors specified in the submission form.

Parameters (optional)

Electrostatic interaction

You can provide the electrostatic interaction energy calculated by the HADDOCK refinement page. If you submit a PDB file obtained directly from the HADDOCK refinement page, PRODIGY-LIG will automatically retrieve the energy from it. Otherwise, the energy can be added manually in the "Electrostatic Energy" box.

If no electrostatic energy is provided , the server will run the "No electrostatics prediction" protocol (see below and in "Methods").

Check also the paragraph below "How do I calculate the electrostatic energy of the complex?"

Job ID

The user can specify a personalized Job ID to identify the run.

Email

If an email is provided, a link with the results will be sent when the job is done. The results will be stored for 2 weeks.

Output

PRODIGY-LIG outputs are displayed online and remain downloadable for 14 days. A link to the online resource is also emailed to the user, if an email is provided.

The results returned by the server include:

1. Predicted value of the binding affinity (ΔG) expressed in kcal mol-1
2. Number of intermolecular atomic contacts (ACs) at the interface between the protein and the ligand within the threshold distance of 10.5 Å. Contacts are divided by atom-type.
3. Downloadable table (.txt) listing the residues in contact within the given threshold
4. Archive file including all the output files

How do I calaculate the electrostatic energy of the complex?

To refine your protein-small ligand complex and obtain the electrostatic energy, go to the HADDOCK refinement web page. Make sure to register for a HADDOCK account in order to submit jobs, use the registration page). In this example, we will refine the protein-ligand complex with PDB code 4JRG.

Note: The blue bars on the server can be folded/unfolded by clicking on the arrow on the right

Step 1 Define a name for your docking run, i.e. 4JRG_refinement
Step 2 Input the protein PDB file, or type the PDB code with the chain identifier. For this unfold the First Molecule menu.
Step 3 Input the ligand PDB file, or type the PDB code with the chain identifier. For this unfold the Second Molecule menu.
Step 4 You are ready to submit!

Once your run has completed you will be presented with a result page showing the cluster statistics and some graphical representation of the data.

In the refinement, there will be only 1 cluster considering that the models represent just a refinement of the bound input structure. The various components of the HADDOCK score are reported on the result web page.

The Electrostatic energy (in this example -94.1) is the energy feature that you are looking for to provide at the PRODIGY-LIG page. Alternatively, you can also download the top refined model from the refinement page (click on Download next to "Nr 1 best structure") and upload it to PRODIGY-LIG. The information about the electrostatic energy is already contained in the file and will be automatically retrieved from our web-server.

How do I use Prodigy-Lig without the electrostatic energy?

PRODIGY-LIG can also be used without the electrostatic energy, providing a prediction with lower confidence (see also "Methods"). However, PRODIGY-LIG has shown better performance when the electrostatic energy was included in the model, therefore we strongly recommend to provide it.

Can I provide the electrostatic energy calculated by other tools than HADDOCK?

No, PRODIGY-LIG is an empirical approach that was trained by using the electrostatic interactions energy calculated by HADDOCK. The prediction formula has been therefore optimized according to this feature.