28.3.12.1. KWIK with a single target

In a normal Grid run with a single Target KWIK can be set to 1, 2, 5, 6, or 7 with the following results (Also see Diagram 5 below):

• It should be noted that a complete grid is normally a rectangular box. Some of the grid points lie in the corners of the box, where they may be far from the Target molecule. However, a full energy calculation is normally made for each of these grid points, although the information may not actually be required by the User for such distant positions. Setting the value KWIK=1 forces the Programme to omit any grid point which is more than 4.0 Angstrom away from the Target. This makes the Programme run more quickly. When KWIK=1 the energy values for the omitted grid points are set to zero, and this can sometimes produce a sharp energy jump at the 4.0 Angstrom cut-off distance. Anomalies may also occur if you are studying a system in which the Probe is always repelled from the Target at every point on the grid. Care must therefore be taken when interpreting results with KWIK=1. Note also that bigger values of directive CLER will automatically be reduced to 4.0 whenever KWIK=1. The value KWIK =1 may not be used if you are studying several Target molecules one after the other as a Set.

• Setting KWIK=2 has the same effect as KWIK=1. In addition, however, the detailed geometry of the Probe group as defined by JTYPE is neglected. JTYPE is reset equal to zero, and this further speeds up the computation (see Diagram 5 below). The value KWIK = 2 may not be used if you are studying several Target molecules one after the other as a Set.

• Setting KWIK=5 has the same effect of KWIK=1, but the process of definition of distant-points has been improved. Therefore, when KWIK=5 the Programme Grid will omit any grid point which is more than 5 Angstrom away from the Target (in the case of Uncharged Probes) or 7 Angstrom (in the case of Charged Probes). The energy values for the omitted grid points are set to zero. Directive KWIK is automatically reset equal to zero when LEAU=3.

• If you set KWIK=6 grid points which are inside the Target are easily found and Programme Grid will omit any grid point which is more than 5.0 Angstrom away from the Target (in the case of Uncharged Probes) or 7 Angstrom (in the case of Charged Probes). The effect of setting KWIK=5 is maintained when setting KWIK=6; thereby, two combined factors speed up the calculations. The energy values for the further omitted grid points are set to the GRID energy cut-off (+5 kcal/mol). KWIK=6 is optimal for macromolecules. Directive KWIK=6 is automatically reset equal to zero when LEAU=3.

• Setting KWIK=7 all the effects of KWIK=5&6 are maintained. In addition, the regular grid cube around the (macro)molecule will be modified into a regular ellipsoid inscribed inside the cube. The eights corners of the cube are held out from the computation. Again, the energy values for the further omitted grid points are set to zero. Directive KWIK is automatically reset equal to zero when LEAU=3.

Note that any value of KWIK>0 may not be used unless Directive MOVE=0. This value of MOVE does not permit atoms of the Target to move in response to the Probe. If they were allowed to move, they might go to places near grid points at which KWIK was going to abort the computation. Important results at such points might then be lost if KWIK>0, and this is why these KWIK values are only allowed with MOVE=0. The value KWIK=5&6 can be used if you are studying several Target molecules one after the other as a Set, since the cage dimensions are not modified. KWIK=7 is optimized for macromolecules, when the active site shape is almost regular and quite-spherical.

28.3.12.2. KWIK with a Set of Targets

The normal input to Programme GRID is a GRINKOUT file previously prepared by Programme GRIN, but an alternative procedure is available. Instead of reading the actual GRINKOUT file into GRID through channel INPT, one can read a list of pointers to several GRINKOUT files which are then processed as a Set. See also Chapter 36. KWIK should not have the values 1 or 2 when this is done, but may be 3, 4, 5, 6, or 7 instead.

• If you set KWIK=3 then the size of the grid will be automatically adjusted so that it surrounds all the Targets in the Set. It is essential for all the Targets to be in the same position and arranged in the same orientation; i.e. they must be more or less superimposed. Then (if KWIK has the value 3) the size of the grid will be increased appropriately, and will take account of the fact that some of the compounds may be charged and therefore have a far-reaching electrostatic field. The final size of the grid will be determined according to the superposition arrangement; the charges of each compound; the dielectric environment; and the charges of the Probes. Directive CLER is without effect. See Diagram 6 below. The value KWIK=3 may not be used if you are studying one single Target in the usual way. It may only be used when you are processing several molecules one after the other as a Set.

• If you set KWIK=4 then the size of the grid will be adjusted as described for KWIK=3 but the output to the GRIDKONT file will either be shortened by the elimination of 'wasted' grid points, or the 'wasted' points will be filled by a predefined "missing value". By a 'wasted' grid point is meant one that is distant from all the Target molecules, in a region of space where the energy is mostly electrostatic, and where that energy is small and only varies gradually from point to point. Exactly what happens is determined by Directive VALU like this:

  • If VALU=0.0 you will get a shorter output, because the 'wasted' points will be eliminated from the output to GRIDKONT. Of course, you will no longer get results for all the points of a complete rectangular grid, and so it may not be easy to display the results graphically.

  • If VALU has a non-zero value you will get the full set of grid points, but all the 'wasted' points will have the numerical value of VALU. For instance if you set VALU=-99.99 then every wasted point will have the value -99.99 and you could write a statistics program that would treat this as a "missing value".

  With KWIK equal to 4 it is again appropriate for the Targets to be superimposed (See Diagram 6 below). The value KWIK=4 may not be used if you are studying one single Target in the usual way. It may only be used when you are processing several molecules one after the other as a Set.

• Setting KWIK=5 has the same effect of KWIK=1, but the process of definition of distant-points has been improved. Therefore, when KWIK=5 the Programme Grid will omit any grid point which is more than 5 Angstrom away from the Target (in the case of Uncharged Probes) or 7 Angstrom (in the case of Charged Probes). The energy values for the omitted grid points are set to zero. Differently from KWIK=1, Directive KWIK=5 works also when BOTX, ... , TOPZ are used to define the grid cage; on the contrary, KWIK is automatically reset equal to zero when LEAU=3.

• If you set KWIK=6 grid points which are inside the Target are easily found and Programme Grid will omit any grid point which is more than 5.0 Angstrom away from the Target (in the case of Uncharged Probes) or 7 Angstrom (in the case of Charged Probes). The effect of setting KWIK=5 is maintained when setting KWIK=6; thereby, two combined factors speed up the calculations. The energy values for the omitted grid points are set to the GRID energy cut-off (+5 kcal/mol). KWIK=6 is optimal for macromolecules. Differently from KWIK=1, Directive KWIK=6 works also when BOTX, ... , TOPZ are used to define the grid cage; on the contrary, KWIK is automatically reset equal to zero when LEAU=3.

• Setting KWIK=7 all the effects of KWIK=5&6 are maintained. In addition, the regular grid cube around the (macro)molecule will be modified into a regular ellipsoid inscribed inside the cube. The eights corners of the cube are held out from the computation. Again, the energy values for the omitted grid points are set to zero. Differently from KWIK=1, Directive KWIK=7 works also when BOTX, ... , TOPZ are used to define the grid cage; on the contrary, KWIK is automatically reset equal to zero when LEAU=3.

The default value KWIK=0 is normally recommended, unless the available time is particularly short, or the processor is rather slow, or the Target is very big, or a Set of Targets is being studied.

Note that KWIK must not have the value 1, 2, 5, 6, nor 7 when a multi-atom Probe is being used.

Default: KWIK=0.

KWIK=0, 1, 2, 5, 6, or 7 for a single Target.

KWIK=0, 3, 4, 5, 6, or 7 for a Set of Targets.

28.3.12.3. Using KWIK with symmetrical targets

Some very big Targets are symmetrical. For example there is a perfect axis of two-fold rotational symmetry down the centre of the haemoglobin tetramer, and there is a receptor cleft precisely on this axis. The receptor has two identical halves, and a Grid map of the receptor region would also be symmetrical. In such a case as this it might only be necessary to compute the map for half of the Target, since the other half of the Grid map would be exactly the same.

One method of restricting the size of the map to fit the required region would be to use the directives TOPX..... BOTX. An alternative approach would be to set directive KWIK=1 (or 2) as described above. A third possibility would be to mark the atoms of the Target structure so that only one half of it would be considered. This last approach is always combined with the use of KWIK.

In practice, one should not study a Set of large Targets, and directive KWIK should be set equal to 1 (or 2) when a single large symmetrical Target is being studied. Exclamation marks should then be used to flag half the atoms of the Target. KWIK will ignore the flagged atoms when it is determining the the 4 Angstrom cut-off. For example, if all the atoms in one half of the haemoglobin tetramer were flagged with an !, then the cut-off would surround the unflagged half. It would extend 4 Angstroms into the flagged moiety but no further, so that the final Grid map would cover rather more than the flagged half of the Target.

The exclamation mark can be used to flag any unwanted part of the Target. For instance the insulin hexamer which may be described as an (alpha/beta) trimer, and one could flag out two-thirds of the structure. The computation would then only be made over one third of the grid points, but the whole hexamer structure would be taken into account while that one third was computed; i.e. the result would be different to that which would have been obtained if a Grid map had been computed for an isolated alpha/beta pair alone, because the isolated pair would have been treated as if the rest of the hexamer was missing altogether.

The exclamation mark should be placed against appropriate atoms in the ALT column of the PDB input file, before running the first Programme GRIN. Alternatively, it may be placed in the corresponding column of the GRINKOUT file. The exclamation mark is a reserved character in this position, with the specific function of controlling the Target area when KWIK has been set.

Note that KWIK must not have the value 1 or 2 when a multi-atom Probe is being used, and the exclamation mark may not be used to flag symmetrical atoms with multi-atom Probes. The interactions between Directives KWIK and LIST are summarised in Diagrams 5 and 6 (see Directive LIST).

28.3.13.1. Bridging waters

In some circumstances a water molecule may act as a Hydrogen Bonding Bridge between Target and Probe. For example, Matthews et al found such a water between a threonine hydroxyl group of dihydrofolate reductase, and an NH2 substituent of its ligand methotrexate. The presence of this water was an essential feature in the binding pocket.

Programme GRID will take explicit account of such a Water molecule when LEAU = 1 The User does not need to provide any water coordinates, although waters may be included in the input PDB and GRINKOUT files if their positions are known. The main GRID computation for Target and Probe is started as usual. Then, if LEAU = 1, GRID will find and take account of any sites for Bridging Waters whilst it does the main computation for the interaction between Target and Probe. A conventional lineprinter output GRIDLONT is produced, showing favoured sites for the Probe as usual. If a Bridging Water contributes to the Probe:Target interaction, then an extra line for that Bridging Water will be printed to the GRIDLONT file. The energy value which is written to the GRID map generated in the GRIDKONT file, will also include an appropriate contribution from the bridging water.

The value LEAU = 1 may only be used with certain Probe Types. The Probe must be able to make at least two hydrogen bonds, one to the Bridging Water molecule and one direct to the Target. Probe Types with more hydrogen bonds can of course be used. The geometry of the Probe must be defined by JTYPE, and so JTYPE = 0 is not an acceptable value.

The value LEAU = 2 may only be used with a few Probe Types. The Probe must be able to make at least three hydrogen bonds, two to the Bridging Water molecules and one direct to the Target. Water, which can make four hydrogen bonds may of course be used. The geometry of the Probe must be defined by JTYPE, and so JTYPE = 0 is not an acceptable value.

If the main Probe is itself water, then two possibilities are considered when LEAU = 2:

• The water forms one hydrogen bond direct to the Target, and one or two to bridging waters.

• The water forms two hydrogen bonds to the Target, and one or two to bridging waters.

Thus, when the main Probe is water, a network of water molecules across the surface of the Target can be considered, each forming sideways hydrogen bonds to its water neighbours.

When Directive MOVE = 1 or 2 or 3 or -1 certain atoms of the Target are allowed to move under the influence of the Probe. In these cases the values LEAU=1 and LEAU=2 must not be used.

28.3.13.2. Competiton by the probe

It sometimes happens that the chosen Probe may interact favourably with the Target at a certain position, but that a water molecule would interact better at the same place. The net interaction energy of that particular Probe at that grid point would then be unfavourable, and Grid will detect regions where this happens if directive LEAU is set as LEAU=3. More detailed information will be found in Chapter 34.

It is important to know if specifically unfavourable places for certain Probes exist on a Target, because they can influence the affinity and selectivity of drug-receptor interactions, and the orientation of a ligand at its receptor site. A well-known example occurs with the enzyme dihydrofolate reductase, which binds the purine ring of its substrate in one orientation, and the corresponding ring of its inhibitor methotrexate (MTX) after an 180 degree ring rotation. The significant structural difference between these two ligands is the replacement of an oxygen atom in the substrate by a nitrogen in MTX, and the inhibitor was originally designed on the assumption that it would bind exactly like the substrate itself.

The bound orientation of MTX was observed many year's ago by X-ray crystallography. This Version of Grid shows (when LEAU has been reset to 3) that an oxygen of the substrate would be precisely located at a place in the enzyme cleft which is maximally unfavourable to oxygen, if the substrate were to bind in the same way as MTX. Indeed it appears as if two backbone carbonyl oxygens of the enzyme are an important structural feature which ensures that the substrate does not bind like MTX, because dihydrofolate reductase can only give a stereochemically correct product if the substrate ring is rotated 180 degrees away from the MTX orientation. See Chapter 60 for more information.

Results such as these (with LEAU=3) suggest that ligand selectivity may be critically influenced by pre-bound water molecules, because water is not easily displaced by polar but inappropriate ligand atoms. One misplaced polar atom such as carbonyl may give such a big positive energy that it forces the ligand into a completely different orientation, and causes several otherwise apparently acceptable hydrogen bonds to be misaligned.

Default: LEAU = 0

Alternative Values: LEAU = 1 or 2 or 3

28.3.14.1. Details about negative values for LENG

Sometimes there are Target atoms which have strong positive interaction energies (indicating repulsion of the Probe), and they will be brought to the top of the table when LENG is negative and the repulsion energy is substantial. This is done because it could be misleading to look at all the attractions and ignore significant repulsions. For instance, the charge on a carbonyl carbon is typically +0.3, and the charge on the carbonyl oxygen -0.3, so that one might be strongly attracted and the other strongly repelled by an anionic Probe. In this situation it would then be misleading to consider the attraction alone.

In other situations there may be lots of Target atoms which show a weak attraction, and lots more which show weak repulsion. These weak energies may not be so interesting, and such atoms will all be consigned towards the end of the table when LENG is negative.

Default: LENG = 30

-251 < LENG < 251

28.3.17.1. Using directive LIST with a single target

When a single Target is being studied in a normal Grid run, an extra record will be prefixed to the start of the computer-readable GRIDKONT file if LIST=1 which is the default value. This record will be absent if LIST=0

The extra record contains additional information which can be used, for example, when a graphics program is being initialised in order to display the Grid map.

28.3.17.2. Using directive LIST with hypermolecules

Directive LIST is provided because it may not be possible to generate the whole GRIDKONT file in a single computer run, when a very large macromolecule is being studied on a slow computer. The individual planes of grid points (i.e. the z-planes; see below) can then be calculated in batches, and finally merged to give the complete GRIDKONT file.

You can achieve this by setting directive LIST=1 when you calculate the first batch of z-planes, and having LIST=0 for the remaining batches. Then, when you merge the output files the extra information will only be provided once at the start of the merged GRIDKONT file.

28.3.17.3. Using directive LIST with a set of targets

The situation is somewhat different when a Set of Targets is being studied, and LIST has its default value of 1. In this case LIST forces the GRIDKONT output to begin with a list of the x,y,z coordinates of the grid points. These are followed by the Grid energies for each Target molecule, one after the other. The energies are printed for all the Targets with the first (or only) Probe, and the whole printout is repeated if there are subsequent Probes. Thus the layout of the GRIDKONT file will be arranged like this:

 The set of coordinates for Probe A 
   Energies for Target 1 with Probe A 
   Energies for Target 2 with Probe A 
   . 
   . 
   Energies for last Target with Probe A 
 The set of coordinates for Probe B 
   Energies for Target 1 with Probe B 
   Energies for Target 2 with Probe B 
   . 
   . 
   Energies for last Target with Probe B 
 The set of coordinates for the last Probe
   Energies for Target 1 with last Probe
   Energies for Target 2 with last Probe 
   . 
   . 
   Energies for last Target with last Probe

You may set LIST = -1 only when studying a Set of Targets. This value of LIST then shortens the GRIDKONT printout. The coordinates and energies are printed as usual for the first Probe, but the coordinates are not repeated before the set of energies generated by the second or later Probes :

 One set of coordinates for all Probes
   Energies for Target 1 with Probe A
   Energies for Target 2 with Probe A
   . 
   .  
   Energies for last Target with Probe A 
   Energies for Target 1 with Probe B 
   Energies for Target 2 with Probe B 
   . 
   . 
   Energies for last Target with Probe B 
   Energies for Target 1 with last Probe 
   Energies for Target 2 with last Probe 
   . 
   . 
   Energies for last Target with last Probe

Note that LIST should not be set equal to zero if several Targets are being studied as a Set. If you do set LIST = 0 for a Set, then the value of LIST will be changed from 0 to 1.

Setting LIST = 2 is only permitted when a Set of Targets are being studied one after the other. The output is arranged as described above for a Set with LIST = 1, but is printed in computer readable binary code. This gives a smaller output file which can be written more quickly, and can be read more quickly by the computer on which it was written. On the other hand it cannot be read by eye, and may not be compatible with other computers.

Setting LIST = -2 is only permitted when a Set of Targets are being studied one after the other. The output is arranged as described above for a Set with LIST = -1 but is printed in computer readable binary code. This gives a smaller output file which can be written more quickly, and can be read more quickly by the computer on which it was written. On the other hand it cannot be read by eye, and may not be compatible with other computers.

28.3.17.4. Subdividing the gridkont file

When a Set is being studied and LIST = 1, then the GRIDKONT file may be relatively large. However it will be easy to sub-divide the file into a number of sections, one for each Probe and all having the same layout of coordinates followed by Grid energies. On the other hand when LIST = -1 the GRIDKONT file will be shorter because the coordinates will not be repeated for each Probe. In this case, however, it may not be so easy to sub-divide the file into similar sections with the same layout in each.

28.3.17.5. Summary for directive LIST

In summary, the available values of LIST are: Default: LIST=1; LIST=0 or 1 for a single Target; LIST=2 or 1 or -1 or -2 for a Set.

>DIAGRAM 5: SUMMARY OF THE EFFECTS OF DIRECTIVES KWIK AND LIST WHEN ONE SINGLE TARGET IS BEING STUDIED BY ITSELF. (NOTE THAT DIFFERENT EFFECTS ARE PRODUCED WHEN A SET OF TARGETS ARE BEING STUDIED ONE AFTER THE OTHER. IN THAT CASE LIST MAY NOT HAVE THE VALUE 0.)

Combinations marked with a star * are NOT recommended.

>DIAGRAM 6: SUMMARY OF THE EFFECTS OF DIRECTIVES KWIK AND LIST WHEN SEVERAL TARGETS ARE BEING STUDIED AS A SET. (NOTE THAT DIFFERENT EFFECTS ARE PRODUCED WHEN ONE SINGLE TARGET IS BEING STUDIED BY ITSELF.)

* If you set LIST = 0 it will be reset to LIST = 1

Combinations marked with a star * are NOT recommended.

28.3.18.1. The regular effects of direct MOVE

When directive MOVE=1 in GRIN, and MOVE=1 or 2 or 3 in GRID, the effects are like this:

• Programme GRIN computes extra information about the Target, and prints this to the GRINKOUT file which becomes the input to GRID.

• Programme GRID assigns each Target atom to one of three main classes: either it is in the Core where it cannot move in response to the Probe; or it can move as part of a flexible chain; or it can move as part of a Bead. See Section 21.4 for more detailed information.

• Programme GRID uses the traditional Grid algorithms to deal with atoms in the Core of the Target. It treats Target atoms and Beads which can move as described in Section 21.4.5.

28.3.18.2. The special value MOVE = -99

A special value MOVE=-99 may be used when Probe Molecules are being studied (See Section 41.5). Grid normally tries to adjust the position of a Probe Molecule in its binding cleft, in order to relax close contacts. This relaxation is recommended, but will not take place if MOVE=-99 in the Grid run.

28.3.19.1. Using NETA on a single target

Programme Grid is normally run on one single job with one single Target. This single Target may be just one molecule, or it may contain several proteins, ligands, counter-ions and solvent molecules making a more complex Target. Alternatively, Grid may be used to study several Targets one after another as a Set. This Set Mode is used, for example, if a set of Grid Maps are required for a set of related drug molecules, so that Statistical analyses can be performed on the individual energy values generated by Grid.

When there is only one Target (Macromolecule Mode; or Small Molecule Mode; or Mixed Mode) the following arrangements for NETA will apply. The arrangements are different when a Set of Targets is being studied (see below).

In order to understand the usefulness of assigning a non-default value to NETA, it is necessary to study the input file GRINKOUT. This file is a list of the ATOMS in the Target, and these ATOMS may be followed by a list of HETATMS and finally a list of computed hydrogen positions. When the Target is a protein or other macromolecule, the HETATMS may include bound water, other solvent molecules, ligands, substrates and so on.

Programme GRID would not normally treat the HETATMS as a part of the macromolecular Target, if ATOMS and HETATMS were both present in the GRINKOUT file; i.e. if Programme GRID was being used in Mixed Mode. In this case the default would be to ignore the HETATMS, and only treat the ATOMS of the macromolecule as the Target. The default value is NETA = 0. This default value is used, so that the HETATMS which represent water molecules will not be considered as part of the macromolecular Target in Mixed Mode. These water molecules may be present in large numbers in the original PDB file for a protein. If the User wishes to treat some of the HETATMs as a part of the Target, then Directive NETA must be assigned an appropriate value as described below.

Macromolecule Mode deals with Targets which contain only ATOMS. Small Molecule Mode deals with Targets which contain only HETATMS. The default value of NETA=0 is also appropriate in both these modes. In this case, however, Programme GRID will automatically change NETA to the correct value (if necessary) like this:

28.3.19.2. Using NETA in macro-molecule mode

In Macromolecule Mode, when there is only one Target which contains ATOMS but no HETATMS, the value of NETA will be zero; ie: no change will be made from the default.

28.3.19.3. Using NETA in small molecule mode

In Small Molecule Mode, with one Target which contains HETATMS but no ATOMS, the value of NETA will be increased to the number of HETATMS in the Target.

28.3.19.4. Using NETA in mixed mode

NETA should be set by the User when Mixed Mode is required; ie: when the Target contains both ATOMS and HETATMS. In this case NETA may be used in one of four different ways if a single Target is being studied. Different arrangements are made if several Targets are being studied together as a Set (see Chapter 36).

Users need only study Method 1 initially. This method is the most straightforward and can often be used. At a later stage most Users also try the other methods, and then choose the approach they prefer. The four methods are:

1. Consider by way of example a macromolecule which has two metal groups firmly embedded in the macromolecular matrix, notwithstanding that the metals are actually hetero groups; i.e. they are HETATMS with Energy Variables from the HET section at the end of datafile GRUB. Assume, for the sake of simplicity, that these two metal groups are the only HETATMS in the Target. By default these two HETATMS would be excluded from the Target because ATOMS are also present (Mixed Mode). However the metal groups are firmly embedded in the protein, and the User might NOT want to treat them as displaceable ligands but rather as a part of the macro- molecule. In this case an instruction would be needed to tell Programme GRID that the two HETATMS should be used in the computation just like normal Target atoms. This would be done by setting NETA=2 It must be emphasised that this first method would not work correctly unless the two metal groups were the only HETATMS in the Target. Moreover they must come at the end of the PDB file, immediately after any ATOMS. (Note, however, that a TER record would be required just before the HETATMS, at the end of the ATOM list, if the ATOMS constituted a protein chain ending with a terminal carboxy group. The TER record would come after the OXT ATOM at the end of the protein chain).

2. Secondly consider a PDB input file containing many ATOMS and many HETATMS. To use NETA effectively in this case it would first be necessary to edit the PDB file so that the wanted HETATMS (i.e. the HETATMS which were to be treated as part of the Target) came IMMEDIATELY after any Target ATOMS. They must be moved to the top of the HETATM list in the original PDB file (This is most easily done with a page editor) before Programme GRIN is used to generate the GRINKOUT file for input to GRID. This GRINKOUT file will start with Target ATOMS which have been assigned Energy Variables from the Recognised Molecules at the start of datafile GRUB. It will then contain the HETATMS which are to be part of the Target, and these will be followed by any other HETATMS in the PDB file. Finally there will be a list of hydrogens. The HETATMS and hydrogens will be marked with a star * or a cross X. When running Programme GRID, directive NETA is the number of HETATMS which will be included with the ATOMS of the Target molecule. These are the Extra Target Atoms, and NETA is the "Number of Extra Target Atoms". In effect, NETA moves the pointer between ATOMS and HETATMS, so that some HETATMS at the top of the HETATM list are reclassified with the ATOMS of the Target, and are therefore used in the GRID computation. It is therefore necessary to set NETA equal to the number of required HETATMS which were moved with the page editor to the top of the HETATM list in the original PDB file. Note that the first method of using NETA, described above, is a special case of this second method. In the first method it is assumed that all the HETATMS are wanted. Another way of including all the HETATMS is to set NETA to a very large value, in which case Programme GRID will reduce it appropriately to include exactly all the HETATMS.

3. The third method is similar to the above, but the page editor is used to remove all the unwanted HETATMS from the PDB file. A new GRINKOUT file is then created with Programme GRIN, and this consists of ATOMS followed by HETATMS and hydrogens. All the remaining HETATMS are wanted in the Target, and the User should set directive NETA to a large positive value. Programme GRID will automatically reduce it to the exact number of HETATMS present.

4. The fourth procedure is to begin with the standard GRINKOUT file, and set NETA to a large value so that all the HETATMS will be included with the ATOMS of the Target. NETA will by default be reduced by Programme GRID to the exact number of HETATMS and a message will be sent to the lineprinter file GRIDLONT at run-time. Before starting the GRID run, however, the cross X is used to exclude any unwanted ATOMS or HETATMS from the input file GRINLOUT. Thus Programme GRID will only use for its Target the remaining ATOMS and HETATMS which are uncrossed. A possible problem with this fourth method, is that the hydrogens bonded to a heavy atom must also be crossed off the GRINKOUT file at the same time as their heavy atom itself. This will usually call for careful book-keeping.

28.3.19.5. Using NETA in set mode

When a set of Targets is being studied all together, one after the other as a Set, the following arrangements for NETA apply. The User should first read the above Sub-Section describing how to use NETA on a single Target, and should also look at the main Section on the use of Grid in order to study a Set of Targets (see Chapter 36). Then look at the following description.

The individual Targets of a Set are normally individual small molecules which are being studied one after the other. Each molecule is composed of HETATMS. The default value of NETA is set equal to Parameter MAXDIM, which is normally 24000. Programme Grid then reduces this default to the appropriate value for each successive Target, and so each Target is processed correctly.

Other Users may wish to study a Set of Targets in which each individual Target contains both ATOMs and HETATMs. For example each Target might be a protein (ATOMs) and the counter-ions (HETATMs) which are required in order to have overall electro-neutrality. Once again the default value of NETA is set equal to MAXDIM, and this is reduced by Programme Grid to the appropriate value for each Target; ie all the counter-ions would be included as part of the Target.

This default arrangement is different from the default in Mixed Mode when a single Target of ATOMs and HETATMs is being studied. This is the difference:

• By default with one Target in Mixed Mode, none of the HETATMs are included with the ATOMs of the Target. The default value is NETA = 0 and this default must be changed if some HETATMs are needed.

• By default in Set Mode, all of the HETATMs are included with the ATOMs of their Target. The default is NETA = 24000 and this default cannot be changed by the User. Thus, as each successive Target of the Set is processed, it will be processed by Programme Grid with all its ATOMs and all its HETATMs. Therefore, if you are working in Set Mode, and only want to have ATOMs in the Target, you must prepare the input file for Programme Grid, so that it does not contain any HETATMs.

The reason for this different default is as follows. New Users should have a simple default in Mixed Mode, so that the receptor cleft of the enzyme which they are studying is not completely blocked by solvent molecules. Their default is therefore NETA = 0, and they can increase this to the appropriate value if they wish to include some counter-ions or other HETATMs as part of the Target in Mixed Mode.

In Set Mode, however, each successive Target might need a different value of NETA. For example, the first protein Target might have six counter-ions and the second might have ten. If the default value was NETA = 0 then the User would have to assign an individual value of NETA for each Target of the Set, and this would be cumbersome. Therefore the default is NETA = 24000 which is automatically reduced to the appropriate number for each successive Target of the Set.

One important difference between the default arrangements should be noted. None of the HETATMs is included by default with the ATOMs of the single Target in Mixed Mode. On the other hand all of the HETATMs are included by default with their Target, as the successive Targets are studied one after the other in Set Mode.

WATER MOLECULES

In some computations the User may be able to delete all water molecules from the PDB input file, before it is submitted to GRIN and GRID. Deleting the waters like this will give faster runs, than setting NETA=0 while leaving the waters as HETATMS in the PDB file. Whether the waters can be omitted of course depends on the objective of the particular computation.

TER RECORDS

When the ATOMS of the Target constitute a protein chain ending with an OXT atom and a carboxy terminal group, there should always be a TER Record after the OXT carboxy oxygen. This is particularly important if HETATMS are also present and Directive NETA is used.

Defaults when there is only one Target:

Defaults when there are several Targets which are being studied one after the other as a Set. The User cannot reset these defaults for NETA:

Limits: 0 <= NETA <= Total number of hetero atoms.

28.3.22.1. HETATM records as POSI directives

Brookhaven Protein Data Bank HETATM records may be used instead of POSI directives. Each HETATM record must exactly comply with the HETATM format as specified by Brookhaven. In particular the 'H' of HETATM must be the first character on the line. A whole list of HETATM records can be included in the command file for GRID, in exactly the same way as a list of POSI records.

This is particularly convenient because many programs give output in PDB format. For instance the output from Programme MINIM is a list of all the favourable minima in a Grid map. If such a map had been generated by a water Probe, the list of minima would be a list of favourable water positions. It could then be used as a POSI list, in order to run GRID and obtain a GRIDLONT lineprinter output table describing the environment of each water position.

28.3.22.2. Output when POSI has been set

 HETATM    1 ATM  MOL  -999      66.000   3.000   7.000 -1.76

28.3.22.3. Generating POSI values

The output from many Programmes is a list of atom positions written in Brookhaven HETATM format, and these can be used as a batch of POSI values. In other circumstances the User may want to write a jiffy program which will generate a list of the POSI positions at which Grid energies are required. For instance he or she might want grid points distributed at spherically or cylindrically defined positions, or over the Van der Waals surface of the Target, instead of an orthogonal Grid.

Any distribution of grid points can be achieved, and pasted as a set of POSItions to the command file for Grid. It should be noted, however, that some graphics programs may not be able to display the results unless the Grid map is conventionally arranged.