Interaction statistics
Home > Interaction statistics
Macromolecules like proteins and nucleic acids orchestrate a vast array of intricate cellular processes that are essential for maintaining biological systems. A central aspect of this coordination is molecular recognition—the precise and specific interactions between molecules, often mediated by non-covalent forces such as hydrogen bonds, van der Waals forces, and electrostatic interactions. Understanding these interactions, down to their atomic-level details, is crucial to unlocking insights into fundamental biological functions and processes.
Protein Data Bank (PDB), the single repository of more than 200,000 structures of macromolecules provides a wealth of 3D structural information of macromolecules and their interactions. While PDB is a repository of macromolecular structures, almost 75% of PDB entries contain one or more small molecules, providing wealth of data for analysis of protein-ligand interactions. The analysis of protein-ligand interactions is particularly important because these interactions govern key biological functions across various systems. For example in enzymes, the interactions between the protein and substrates determine the biochemcial reaction efficiency and specificity. Similarly, interactions between ligands and receptor proteins control signaling pathways.
To make the structural analysisof protein-ligand interactions easier for the researchers, the interactions between ligands and biological assembly of all the deposited PDB entries are calculated using PDBe Arpeggio,and made available through PDBe entry pages and PDBe API. This information is quite useful if we are interested in instancesof ligands within a PDB entry. But, what if we are interested in understanding the type of interactions or the type of amino acids most frequently interacting with a ligand? In order to answer such type of questions from ligands' perspective, we have aggregated the interaction data for all the ligands from the PDB entries they are bound, and presented in the Interaction statistics section of PDBe-KB Ligands.
Interaction statistics of imatininb (STI)
The interaction statistics shows a summarised view of aggregated protein-ligand interactions from the PDB archive. The visualisation includes mainly two interactive components
The heat map shows Ligand atom interaction ratio and Ligand-amino acid interaction ratio calculated as below:
$$\text{Relative interaction frequency of ligand atom} = \frac {\text {Number of interactions by a ligand atom}}{\text {Number of interactions by all ligand atoms}} \times 100$$ $$\text{Relative interaction frequency of ligand-amino acid pair} = \frac{\text{Number of interactions between a ligand-amino acid pair}}{\text{Number of interactions by the amino acid}} \times 100$$
- Zoom-in/out - Use the top tracks to zoom in and view ligand atom labels
- Tooltips - Hover over cells to view data in a tooltip. Clicking a cell pins the tooltip; you can pin up to two tooltips to compare values
- Grouping and Sorting - Amino acids are displayed with distinct icons based on their properties. You can group the heat map by these properties or sort it by interaction values
- Filtering: Use the multi-select boxes in the left panel to filter interaction types. By default, all types are included, but selecting specific types updates the heat map accordingly
The 2D view shows interactions per ligand atom by overlaying circles on each ligand atom in its 2D structure.
- Circle size and color: These indicate the interaction value per ligand atom, with larger and darker circles showing higher values.
- Tooltips: Hover over circles to see a tooltip with the atom name and value in the bottom-left corner.
- Atom names: Toggle atom labels using the View Atom Names button.