This web service is intended for the structural prediction of protein-protein interactions using an improved version of our Fast Rotational DOCKing method*. Given the 3D coordinates of two interacting proteins (conforming PDB format) the server efficiently generates many potential predictions of how they could interact. Once your job is completed you can interactively check protein-protein models and add experimental constraints to refine the docking.
A standard-size case usually takes only a couple of minutes to complete the docking depending of the size and computational resources available. The users can choose between scoring weighting schemes optimized for three different interaction types: Enzyme/Substrate, Antibody/Antigen, and Others to slightly improve the success rates. Please keep track of your jod ID number so you can have access to your results from the Results tab. Alternatively you can also choose to be notified by email.
New version 2020 (v3.12)
| *If you would like to refer our work use: | ||
| Server/Improved method | E. Ramírez-Aportela, J.R. López-Blanco, and P. Chacón (2016). FRODOCK 2.0: Fast Protein-Protein docking server. Bioinformatics, 32(15), 2386-2388.   (Please, use this reference) |
|
| Original method | J.I. Garzón, J.R. López-Blanco, C. Pons, J. Kovacs, R. Abagyan, J. Fernández-Recio, and P. Chacón (2009). FRODOCK: a new approach for fast rotational protein-protein docking. Bioinformatics, 25, 2544-2551. |
|
Here you can find the best docking results found in Benchmark 4.0. If you click on any of the case links, an interactive Jmol showing the first acceptable solution and the bound ligand pose will appear (note that these results are from an older version, the current Frodock server is likely to yield better results).
Antigen/Antibody (1FSK) 
|
Enzyme/Substrate (1JTG) 
|
Others (1GPW) 
|
Here you can find all the best docking results found in Benchmark 4.0.
|
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
- Step 1.
Input data - Step 2.
Submit job - Step 3.
Queue - Step 4.
Inspect results - Step 5.
Download results - Step 6. (optional)
Distance constraints - FAQ
Troubleshooting
First, upload the input coordinates of both ligand (smallest) and receptor (biggest) proteins in strict PDB format. Be aware of missing atoms, alternative conformations, bad placed atoms, and a long etc. that can eventually jeopardize your results. Use your favorite PDB checker (e.g. Molprobity) to anticipate and fix any potential PDB error. You can optionally introduce an email address to receive a direct link to access your results later.
Alternatively, depending of the nature of your interaction you could choose between Enzyme/Substrate, Antibody/Antigen, and Others to slightly improve the success rates. Otherwise, Unknown interaction type will be assumed.
Upon pressing the "Submit Job" button your job will be queued in our system and you will be redirected to the Queue tab. A standard-sized docking usually takes less than two minutes to complete depending on the size and computational resources available.
Here all jobs submited to Frodock Server are listed. Your jobs are shown in darker colors whereas those submitted by others appear in lighter colors. You can check server usage and whether any queued job is running ("r" status, green) or queued ("qw" status, orange). In case any of your jobs has been queued ("qw" status) it will run as soon as computational resources become available.
Once your job is completed you can interactively check protein-protein models and add experimental constraints to refine the docking. Results can be interactively visualized in 3D with JSmol. Use mouse controls to readily explore the solutions by visualizing multiple docking poses at the same time and toggling them between different molecular representations and colors. The user can choose between three different interfaces: Java, HTML5 and WebGL. The HTML5-javascript interface is activated by default. To activate the JAVA, first you must enable it in your web browser (see details) and add frodock.chaconlab.org to the exception sites in the Java panel (see details)
At the end of the page you can find the corresponding links to download all the files generated by the server.
In case you have distance constraints available, i.e. any biochemical or biophysical data indicating the approximate distance between two residues in the complex, it can be specified to select only the compatible poses and thus improve the ranking of the best solution. Please, expand the "Distance Constraints" panel and select two residues (one in the receptor and the other in the ligand), introduce the distance threshold (in Angstroms) in the input box, and click the "Add" button. You can specify up to ten distance constraints online. Finally, upon clicking on the "Apply constraints" button, a selector for the filtered solutions will appear in the "Load Solutions" panel. You can try any of the following constraints as practice.
| Distance constraints examples | ||||||
| 1NCA | Antigen/Antibody | 9 | 1 | ASN-102 | THR-319 | 5.0 |
| 1N8O | Enzyme/Substrate | 3 | 1 | MET-184 | MET-80 | 5.0 |
| 1ZHI | Others | 9 | 1 | SER-113 | PRO-9 | 5.0 |
| 1GCQ | Others | 33 | 1 | THR-205 | PRO-2 | 5.0 |
-
What are the common mistakes?
Format errors are behind of FRODOCKś malfunction in 99% of the cases. Be aware of missing atoms, alternative conformations, bad placed atoms, bad names, and a long etc. that can eventually jeopardize your results. Use your favorite PDB checker (e.g. Molprobity) to anticipate and fix any potential PDB error. Ideally, the input structures should be parsed to check errors and minimized with an external program.
-
What are the differences in the docking algorithm and/or scoring function for different Type of interaction?
There are none algorithm differences. Depending of the protein interaction, the users can choose between scoring weighting schemes optimized for three different interaction types: Enzyme/Substrate, Antibody/Antigen, and Others to slightly improve the success rates. The normalized weighting factors for Wvdw, WEle, WSAS and, Wpair are 1.0, 0.3, 0.5,15.0 for enzyme substrate cases, 1.0, 0.6, 0.5, 15.0 for antigen-antibody cases and 1.0, 1.0, 0.5, 15.0 for Others cases, respectively. And the default parameters (unknown interaction selection) are 1.0, 0.5, 0.5, 15.0. It is important to mention that compared with the addition of the pairwise potential the gain obtained in terms of docking success rates was quite reduced. The biggest improvement (~5%) was found by up weighted the electrostatic term in test cases classified as “Others”.
-
Do the users need to run many docking jobs with different random orientations?
Note that the reported success rates (see ROC plots in benchmark tab) are calculated with 50 docking orientations per test case. Thus, contrary to other methods, the reported success rates give a more realistic estimation of the performance as it includes the docking dependence from initial positions. Nevertheless, FRODOCK is quite robust and when is able to identify the correct pose it does in all tested rotations. However, the exact rank of best prediction found varies; for example, for the cases in where FRODOCK found acceptable solution within 2000 positions (150 cases, success rate of 84%) we found on average a variation of 100 positions. This variation logically depends on the test case and the FRODOCK’s relative sensitivity. When FRODOCK identifies the correct pose in the top 100 the average rank variation is reduced to 15 positions. Despite of this variation and given the representativeness of the benchmark, we expect success rates as displayed in in the ROC curves in benchmark tab independently of the initial receptor and ligand orientation. Therefore, the users do not need to run many docking jobs with different random orientations.
-
¿Visualization problems?
3D display is based on JSmol which is very stable and compatible with all browsers. However, problems may occur. For example, reported problems with the display of the hot spots are likely related with older version of JSmol installed in the browser cache. Please be sure that the JSmol correspond to 14.40_2015.12 ( JSmol menu >> About). If not, the solution will be to clear or empty the browser's cache