14.1.1. Hydrogen bond geometry - lysine as an example

Remember that N1IN is the final variable appended from Datafile GRUB to each atom or hetatm in the Target. It is an integer which specifies the hydrogen bonding Type. The Type defines:

1. The number of hydrogen bonds donated by an atom

2. The number of hydrogen bonds accepted by an atom

3. The geometry of the hydrogens bonded to an atom, and the geometry of the hydrogen bonds which it makes.

4. The motions (if any) of the hydrogen bonds made by an atom.

Interested Users may want to study the following example. Consider the hydrogen bonds donated by the amino group at the end of a lysine side chain in a Target protein. This is an sp³ cationic nitrogen bearing three hydrogen atoms each of which can in principle donate one hydrogen bond. Such geometry is defined as Type 3 in datafile GRUB, and the value of N1IN is therefore 3.

N1IN is always an integer variable defining the "Type" of hydrogen geometry at a heavy ATOM or HETATM in the Target molecule. If the atom can donate hydrogen bonds, N1IN determines the hydrogen bond "Type". Programme GRIN will have already assigned an index number I to the particular heavy atom of the Target which is being studied. This index defines the heavy atom's position in the connectivity matrix as described above.

In the present case index I will refer to our particular cationic nitrogen in our particular lysine residue in our particular protein Target molecule. This nitrogen must have the atom name NZ and the corresponding atom name in datafile GRUB is also NZ, which appears in the section of GRUB devoted to LYS (Lysine). The value of N1IN(I) for our atom is the last number on this line in GRUB, and of course it is the number 3 because this atom has Type 3 hydrogen geometry.

The positions of the three hydrogen atoms bonded to our nitrogen depend on the position of the nitrogen itself, and on the positions of neighbouring atoms in the lysine side chain. Some more integer variables N3(I), N4(I) and N5(I) are therefore needed as pointers to these neighbouring atoms. The suffix (I) indicates that these values refer explicitly to the neighbours of our particular nitrogen atom I.

Now consider the structure of the lysine side chain thus:

In this particular case we need only two pointers N3(I) and N4(I) and pointer N5(I) is redundant. N3(I) is used to specify atom CE, so that the direction of the vector from CE to NZ is defined. N4(I) specifies CD, so that the hydrogen atoms bonded to NZ can be located in positions which do not eclipse atom CD. Thus the knowledge that atom I is Type 3 has the following effects. It:

1. Informs Programme GRIN that two pointers to nearby atoms are necessary and that the third is redundant;

2. Directs GRIN to calculate values for these pointers;

3. Informs GRIN that three tetrahedrally arranged hydrogens are needed;

4. Directs GRIN to use the pointers in order to calculate the three hydrogen positions, and add them to the end of the GRINKOUT coordinate list;

5. Directs GRIN to output the two pointers to output file GRINKOUT, so they are available to the next Programme GRID;

6. Directs GRIN that no hydrogen bonds are accepted by this atom, hetatm or extended atom.

It should be noted that this interpretation of N3(I) and N4(I) is specific for Type 3 geometry; the pointers would have different meanings with other geometry Types.

The actual numerical values of the pointers will now be considered. They must point to nearby lines in the GRINKOUT file in order to identify the atoms which determine the hydrogen geometry. In the present example the explicit values of N3(I) and N4(I) are decided by reference to the sequence of ATOM records for the lysine side chain. This is a "Recognised Molecule", and the sequence is therefore defined in datafile GRUB as being:

1   N        In this case the tetrahedrally arranged hydrogen
2   CA       atoms (Type 3) are bonded to atom NZ which
3   C        occurs on row 9 of this GRUB list.  Their
4   O        geometry is determined, as explained above, by
5   CB       the relative positions of the two preceding
6   CG       atoms in the side chain.   These are atoms CE
7   CD       and CD which occur in the GRUB list on rows 8
8   CE       and 7.  Hence N3IN will have the value -1 which
9   NZ       points to CE on the row before NZ in the GRUB
10  OXT      list, and N4IN will be -2 pointing two rows before NZ.  i.e:-

N3IN = 8-9 = -1     and    N4IN = 7-9 = -2

It is essential to distinguish between the geometrical structure of the molecule (which determines what heavy atoms are bonded to the atom which bears the hydrogen), and the GRUB listing which is used in order to interpret N3IN, N4IN and N5IN. In some cases two atoms may be directly bonded together in space, but their line entries may be separated in the standard ATOM sequence of GRUB.