Cite CaviDB

CaviDB: a database of proteins cavities features for structure-based drug design

Ana Julia Velez Rueda, Franco Leonardo Bulgarelli, Nicolás Palopoli, Gustavo Parisi

CaviDB Overview

The first step to running CaviDB is providing a valid PDB ID, UniProt ID, UniProt Entry name or Alphafold ID and the web server will automatically load all the chains related to the search and their general data including their length, number of cavities predicted, and relevant cross-references identifiers.

The features obtained for each entry are organized in two main sections describing the general cavities descriptors, including an interactive display for cavities visualization, the inter-cavities contacts, and the activated residues per cavity, and the global protein descriptors.

Cavities Descriptors

All the cavities predicted for each entry can be visualized interactively, by selecting them in the cavities buttons. They will be highlighted in colors both, in the structure, and also in the sequence.

Druggable cavities (with a Druggability score > 0.5) are tagged with an icon , and can be selected or unselected with a button located before the sequence visualization.

The primary structure of each entry can be displayed in two different representations: a letter soup, and periodic table-like visualization.

When selecting the # button, each residue is represented in boxes containing information about the residue type and its position in its center, the cavities in which it is part in the top left margin, and the number of contacts it has in the top right margin.

By default the # button is unselected, so the sequence is shown as a letter soup, in which the residues shows the information about its position and cavities when making hover in them.

All the descriptors calculated by ,Fpocket are shown in tables and star plots, for their easy comparission.

Intercavities contacts

Intercavities contacts heatmap and network plot, with the druggable cavities highlighted with a black border are provided to users.

The residues linking the cavities and the binding energies corresponding to each contact is shown when making hover in the contacts matrices. All the contacts types predicted by Propka are shown in different contacts matrices: Side chain hydrogen bonds (SCHBOND), Backbone hydrogen bonds (BBHBOND), and Coulombic bounds.

Activated cavities

Information about the activated cavities and Net pKa shift per cavity, are displayed in different plots, colored by cavities colors.

CaviDB Advanced Search

There are many different options for searching CaviDB Entries and pre-filtering the results. It is possible to specify an Uniprot ID using the tag 'uniprot:'

When combining the tag 'uniprot:' with the suffix 'AF', then only the Alphafold conformers of the required protein will be shown. By using the tag 'conf' and providing CaviDB with two specific PDB or Alphafold ids separated by a |, can be retrieved these two specific entries.

The two conformers can be selected for the comparison, and by clicking on the 'compare' button the difference between both will be renderized. The left plots and cavities selectors corresponds to the first conformer and the right to the second one.

By clicking on the cavities selector from each conformer, the cavities will be colored in the sequence. Find more search examples in the following table:

Descriptors detail

Cavities descriptors

For more information check the CIDER and Fpocket documentation

Protein global descriptors

For more information check the CIDER, Modlamp, and Propka documentation

Downloads

CaviDB provides Fpocket outputs for each cavity detected per entry, also the global protein features can be downloaded.