Data

   MSDsite allows access to the ligand environment in a particular PDB entry and contains coordinates with the respect to this entry. The derived bound molecule environments include a consideration of the symmetry in PDB X-ray entries. All possible symmetry and translations are considered to list all possible protein ligand contacts. Water coordinates are treated in a similar manner.

Bonds

The Inter atomic interactions stored in PDBe have the classifications:
• Covalent bond
• Ionic bond

     For metal atom interactions the classification as either a covalent or ionic bond has been considered as a probability between the likelihood as whether metal atom interacts more like covalent or ionic coordination using the following:

  For the elements: F Cl Br I At :
  1. If the atom is a PDB HETNAM chemical group, and is not an ion, it can form one covalent bond.
  2. If the atom is a PDB HETNAM chemical group, and is an ion, it cannot form covalent bonds.
  3. If the atom is a member of a PDB HETNAM group that contains other atoms, it cannot form covalent bonds with any atom of another group.

     The following metals are not expected to form bonds to carbon but are allowed O and N coordination, except for organometallic clusters and carbon monoxide: Li Na K Rb Cs Fr Be Mg Ca St Ba Ra La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Ac Th Pa U

     The following elements may form bonds to C and S : As Sb Bi Sn Pb Al Ga In Tl

     For the following elements a rule of thumb (see below) was used to distinguish between covalent and ionic interactions: Zn Cd Hg Sc Y Lu

  Rule of thumb:
  • If distance between two atoms is within covalent bond length then bond is covalent coordination.
  • If distance between two atoms is within (ionic + covalent) / 2 then bond is covalent coordination.
  • Otherwise if distance is within ionic bond length then it is ionic coordination bond.

     For the following elements, the rule of thumb was applied, except for Carbon which was always considered as forming a covalent coordination bond: Ti V Cr Mn Fe Co Ni Zr Nb Mo Tc Ru Rh Pd Ag Hf Ta W Re Os Ir Pt Au.

     B is always covalent in PDB.

• Disulfide bond
• Hydrogen bond
  We consider two types of Hydrogen bonds based on an algorithm/approach.
  1. Hydrogen bonds are calculated using hybridization, energy type and binding geometry of potential donor-acceptor pairby the methods described in HBExplore.
  2. Hydrogen bonds that are calculated using bndlist library.

     Hydrogen bond assignment is further processed where if the first approach is used and if it did not recognize a pair of atoms as potential donor-acceptor then the second algorithm is applied.

• Van Der Waals bond
• Planes
  We consider two types of plane, planar chemical entities and all rings that satisfy a set of rules are treated as planar groups.
  1. Merged full set of smallest rings.

        For the rings set the definition from Oelib is used (build a minimal spanning tree).

        Each residue is examined in terms of containing rings by first finding the full set of minimal rings. The rings having two or more atoms in common are merged. The result is always the desired ring itself.

     A plane is built from the rings found according to the following rules:
     • Validate ring
       • If ring has 8 atoms or more then it is eliminated
       • If ring has elements B P S then it is eliminated
     • At this stage we assume that the rings are valid and we cleanup the set of atoms that define the rings set to form a plane.
       • Eliminate bridge atoms
       • Eliminate metal atoms
     • If plane has more then three atoms then store it in MSD.
  2. Planar chemical entities or sub-structures of 4 and more atoms.

     For example the carboxylate end of a glutamic acid side chain.

     To define such sub-structures Vega atom topology typing was used

     The following rules are applied:
     • Plane must consist of C-3#0, N-2#0, O-1#0 atoms with Vega type, as a terminate C-4#0, N-300, O-2#0 are appropriate, all the other atoms are eliminated. C-3#0 means that general element is Carbon, the atom is covalently bonded to other 3 atoms in the residue, it can be in a ring. The last one digit says whether the ring is aromatic or not, if it is 0 then aromatic ring is not appropriate (planes from aromatic rings was built on the step one - planes from rings, see above).
     • Atoms must be connected and be in a geometrical plane, where atom deviations from the plane are within 0.1 Angstroms.
     • The number of atoms in a plane must be greater than 3.

     Plane-Plane interactions are stored where the distant between the centres of gravity are within 6 Angstroms, the the distance and normal vectors to the planes are stored.

• Non bonding interactions

  A Non bonding type is assigned to interaction between two atoms that are within 4 Angstroms and which bond type has not been classified.

     The following figure shows relations between database tables for storing interactions between small molecules and macromolecules:

Geometry

Sequences and ProSite patterns

Interactive sortable list of PDB entries

The list presents description of PDB entries and has next functionality:
1. Link to details pages for an entry

   Sequence view

   presents parts of chain sequence where it matches ProSite patterns or searched pattern or interacts with a ligand.

   Ligands view

   presents ligands and interaction with their environment. ProSite or searched pattern is presented as well where matched part of the sequence interacts with a ligand.

   Atomic interaction view

   presents interaction on atomic level- bond type, distance, angle between planes and angles for a single atom ligand.

2. Links to 3D visualization tools
3. Link to integrated resource - atlas pages (MSDlite, MSDPro)
4. List can be listed and sorted interactively.

Interactive statistics charts

Ligand binding statistics

It provides a way to view statistical information regarding the interaction of particular ligand(s) with respect to residues that form an active site environment. The environment is a set of neighbour residues that interact with the ligand through a number of interaction types. It is possible to view two sets of ligand information on the same returned graph to allow comparison of the environment of these ligands.

Environment binding statistics

It provides a way to view statistical information regarding the environment of a ligand with respect to ligands. It is therefore possible to define queries based on residue type and find which ligands are bound to this environment.

Atomic bonds statistics

A search is available to present and compare statistical information for a pair ligand - macromolecule residue interaction on atomic level.

Sequence patterns binding statistics

It provides a way to search PDBe for statistical information regarding the pattern interaction with a ligand with respect to ligands. It is therefore possible to define queries based on pattern and find which ligands are bound to this pattern.

Interactive list of ligands from different PDB entries

The order of this page is to provide functionality for 3D alignment of ligands with their environment. Detailed pages for a PDB entry have links to this page so any ligand can be submitted here and then its geometry, environment and interacting patterns can be compared with another one.

The interface has four alignment options:
• Alignment by ligand
• Alignment by environment
• Alignment by pattern
• Alignment by active-site residues