Input options:

The docked poses can be submitted as a single zipped file. If the PDB file is an oligomeric complex, only two chains can be considered at a time and the chain identifier needs to be mentioned, for example: as 'A,B' or 'C,D'. For calculation of interface surface area for the interface formed, user also needs to upload the individual files of 2 protein chains in PDB format, which were employed to perform docking.

The docked poses uploaded for performing the scoring should be in PDB format and must contain "ATOM" records for both the interacting subunits in the same file separated by a "TER" record. Chain IDs are added by the server to the uploaded input files if the chain ID fields are empty and the coordinates of two chains are separated by a "TER" record. If the files uploaded are not in PDB format, the computation stops with an eror message.

Also, if the coordinates for both the interacting proteins are not provided the computation terminates with an error message.
User can input his/her email address and obtain the URL for result page on their email address provided.

Sample run:

There is an option to perform sample run, where the sample dataset consists of five docked poses (sample.zip). Once the run is executed, a list of score for each docked pose is provided where the docked poses are arranged in the descending order of normalized weighted score.

The list can however be sorted according to parameter of interest to the user.

Parameters used in Scoring:

There are several interface parameters being employed in order to rank the docked poses:

  1. Surface Area
  2. Conservation
  3. Hydrophobicity
  4. Spatial clustering
  5. Short contacts
  6. Presence of positively charged residues at the interface:

Conservation at the interface:

DockScore used to obtain the conserved residues in the two interacting chains from ConSurf web-server. However, while implementing DockScore in the form of web-server we have made DockScore independent of ConSurf and the conservation scores per residue are calculated as follows.
The interacting protein chains were used as a query to perform PSI-BLAST in order to collect homologues against SWISSPROT and multiple sequence alignment is build-using CLUSTALW. Conservation scores per residue are evaluated using MOTIFS (R. Sowdhamini). By default, cut-off score of 60 is used to determine the conserved residues. However, user can select this score cut-off to 60 if alignment is built using close homologues and 40 if alignment includes distantly related members.
User can select the database and parameters to be used for the collection of homologues. By default, the searches are performed using PSI-BLAST with one iteration and Evalue 10-5 and database used is SwissProt.
Also, there is an option provided to upload the pre-computed alignment files in PIR format for both the chains. The user uploaded alignment files should contain the protein sequence of the corresponding interacting chain.

Output/Scores obtained after ranking:

The list can however be sorted according to parameter of interest/score to the user.

  1. Score according to each parameter:
  2. These are six different scores obtained for each pose.

  3. Average score:
  4. This is the average of all individual parameters' score

  5. Normalized average score:
  6. This is calculated by considering the weights assigned to each parameter using the training dataset. It is the normalized average weighted score.

The user can also visualize the top-five poses according to the normalized weighted average score. The interface from the individual chains are also.

User can click on the pose ids and go to the page where they can access the individual scores, interface residues from both the chains and view the pose using Jmol.

Download options:

The following files are also available for download:-

  • Interface files for each pose
  • List of scores for all pose ids in csv format