21.3.1. The format of hydrogen records

The User may study the following Section if he or she wishes to interpret the hydrogen records which appear at the end of output file GRINKOUT. However, these records are generated automatically by Programme GRIN, and are used automatically by the next Programme GRID. Users may therefore prefer to skip this Section altogether. They need only remember that HETATM names must be chosen correctly from the last HET section of datafile GRUB, when small molecules are being studied.

The hydrogen records in file GRINKOUT are each printed by Programme GRIN on a single line with format:

     WRITE (KOUT,230) I,NPDB(I),ATM(I),ALT(I),    +ISUB(I),NRES(I),INSERT(I),X(I),Y(I),Z(I)
 230 FORMAT (I5,'*',I5,A5,A1,'HYD',1X,A1,I4,A1,3X,3F8.3)

or if Lennard-Jones variables and electrostatic charges have been assigned to the hydrogen atoms:

     WRITE (KOUT,240) I,NPDB(I),ATM(I),ALT(I),ISUB(I),
    +NRES(I),INSERT(I),X(I),Y(I),Z(I),VDHY,EFHY,ALHY,QQHY
 240 FORMAT (I5,'*',I5,A5,A1,'HYD',1X,A1,I4,A1,3X,3F8.3,
    +12X,F5.2,F5.1,F5.2,F7.3))

All these hydrogen records must and will be printed together after the Target ATOMS and HETATMS at the end of the GRINKOUT file.

21.3.2. Hydrogens whose positions have been computed

All the hydrogen records for GRINKOUT can be generated automatically by Programme GRIN. The process is transparent to the User who must, however, ensure that each ATOM or HETATM is assigned the correct Type. Atom Types are listed and defined above (see under Hydrogen Bond Type). It is the Type which determines the Case and the Case which defines the hydrogen geometry:

Note on Hydrogen-bonding types

The Type of a Target ATOM in a Recognised Molecule is determined by the corresponding ATOM record in datafile GRUB. This record will be listed in the section of GRUB which is devoted to the particular Recognised Molecule.

The Type of a Target ATOM in an Unknown Molecule is determined by the first ATOM record in GRUB which has the same ATOM name, irrespective of the molecule name. This is a default procedure.

The Type of a HETATM is determined by the row (in the final HET Section of datafile GRUB) which has the same HETATM name as the HETATM of the Target. Examination of the HET Section of GRUB shows, for example, that C3 is carbon Type 30. It is the carbon of an aliphatic methyl group, and therefore requires three hydrogens. It is not the complete extended atom comprising one carbon and three hydrogens. If the User wishes, he or she may supply three input records for the three hydrogens in his or her PDB input file. If these hydrogen records are not supplied by the User, Programme GRIN will compute appropriate hydrogen coordinates and include them in the GRINKOUT file as required for a Type 30 carbon HETATM.

The Type of a single-atom Probe is determined by its Symbol, or by the JTYPE value in its list of Energy Variables. The concept of Hydrogen-Bonding Type is not used for multi-atom Probes which have more complex hydrogen-bonding properties .

Programme GRIN places all the calculated hydrogen records immediately after the Target ATOMS and HETATMS in file GRINKOUT. They must be in the correct sequence in order to ensure that the connectivity matrix is not corrupted. Programme GRIN ensures that this sequence is correctly established, and the User must not edit the GRINKOUT file and thus change the original sequence.

For every hydrogen record in GRINKOUT, variable I is the regular line count as usual. This count progresses consecutively from the ATOMS to the HETATMS in GRINKOUT, and then to the HYDROGEN ATOMS line by line. Variable I is the first number on the line.

Note, however, that variable NPDB(I) for each hydrogen atom is the line count I of the heavy atom to which the hydrogen is bonded. This number is used by Programme GRID as a pointer from the hydrogen to its bonded heavy atom.

The molecule name of hydrogen records must always be HYD, and the line count of the hydrogen record itself must always be followed by a * as shown in the above format statements.

21.3.3. Hydrogens whose positions have been observed

When small molecules are studied, the coordinates for the hydrogen atoms may well be supplied with the coordinates for the other heavier atoms. The hydrogens will be in the PDB input file, and must be HETATMS with the name ' H ' Programme GREAT will deal with this name automatically. Programme GRIN will then read the hydrogen coordinates and will initially accept them as regular HETATMS.

At a later stage the Programme will compute coordinates for all the hydrogens as described above. These computed HYD records will be added to the end of the output file GRINKOUT where they will be marked with a star * They will have HYD as their molecule name, and will be given an ATOM name which includes the name of the heavy atom to which they are bonded.

At this point in the GRIN computation many hydrogen atoms may well appear twice in GRINKOUT, once as a regular hetatm and once as a HYD record calculated by GRIN. It is therefore necessary for the Programme to compare the two positions for each hydrogen atom. If the positions are mutually compatible Programme GRIN adjusts its computed coordinates to agree with the originally reported values, and deletes the original record from the output file GRINKOUT. It does this by marking the original record with a cross X (see here for cross X) which ensures that Programme GRID will ignore that line in the file. However both records remain on file in GRINKOUT, so that the User can see exactly what has happened.

If the observed and computed positions for a hydrogen atom are NOT mutually compatible, then Programme GRIN prints a warning to the lineprinter file GRINLOUT. This most often happens if the hydrogen bonding Type (N1IN) of the heavy atom has been inappropriately assigned.

21.3.4. Energy variables for hydrogen atoms

21.4.1. The conformationally flexible model

Grid Probes are not normally isotropic, and they can be rotated by the Grid Programme so that they will make the most favourable interactions with the Target. This is done again and again at one grid point after another, but the heavy atoms of the Target molecule always remained fixed when earlier Versions (before Version 15) of Programme Grid were used. The treatment of the Target and the treatment of the Probe were therefore dissimilar, because the Probe would respond to the environment created by the presence of the Target, but the Target would not make an equivalent response to the Probe.

In fact the movements of any molecule are always influenced by its neighbours, and the Target should respond to the arrival of the Probe. This is what happens for instance, when the folding of a protein is controlled by a chaperone, and the response of Target to Probe is somewhat analogous to protein folding. This response is envisaged as a three-stage process in the conformationally flexible Grid model, and it may be illustrated by considering the approach of a carboxy oxygen Probe (O::) to the side chain of a superficial lysine residue:

• There will of course be an electrostatic attraction between the anionic O:: Probe and the cationic NH3 group at the end of the lysine side-chain. In fact this will be the only significant interaction when the Probe is a long way away, and the NH3 group of lysine will therefore tend move towards the Probe until it is as close as possible to the grid point. This movement will naturally be limited by the chain of methylene groups linking the NH3 group to the backbone of the protein, and the electrostatic interaction will be most favourable when the methylene chain of lysine is arranged all-trans like this:

  

  Only one all-trans arrangement is possible and this is entropically unfavourable. However, it is possible to calculate the geometry quite accurately and estimate the entropic penalty.

• The chain of lysine methylene groups can be twisted in several different ways when the grid point is appreciably nearer, and the Probe is approaching its optimum position. The overall interaction will then be entropically more favourable because there are alternative methylene arrangements.  Furthermore the electrostatic interactions will also be favoured because the two groups (NH3 and O::) can arrange themselves at an optimal distance from each other, and hydrogen bonds may be formed.  The precise geometry of the methylene groups may not be critically important at this stage, but it is again possible to calculate a favourable geometry and estimate the entropic component.

• The optimal position will finally occur when the O::  and NH3 groups are well arranged relative to each other, and when they are also well positioned in relation to the protein backbone and the rest of the Target, and when the whole methylene chain has also found a good location.  One can represent the situation like this:

  =============| __ |=======PROTEIN BACKBONE========== \___/ \ O:: <-- Probe at the NH3 best grid point

• The precise positions of all the flexible Target atoms may now be critically important; there may only be one ideal arrangement which is not easily calculated; and the entropic situation may again be unfavourable. Under these circumstances it is by no means easy to make detailed computations.

It must be emphasised that this three-stage model is only a model (The real-life process is presumably very much more complicated). However, stages 1 and 2 are relatively easy to compute on the basis of this model, while stage 3 is much more difficult. We hope this Version of Programme Grid does a reasonable job with the first two stages, and that it does not do a very bad job with stage three.

21.4.2. Using the conformationally flexible model

Programme GRIN is used first as usual, and the current Version of GRIN normally prepares a "Core Plus Beads Plus Side-Chain" model of the Target. Each atom of the protein is considered, and is either assigned to an "Inflexible Core" of the molecule, or to a Bead, or to a "Flexible Side-Chain". For example, a superficial arginine residue would probably have its backbone atoms (N, CA, C and O) and its beta methylene group (CB) assigned to the "Core" of the protein, and the NE, CZ, NH1 and NH2 groups assigned as a "Bead", while the CG and CD methylene groups would be in the "Flexible Side-Chain". In this case the CB atom would be called the "Linker" because this is the point at which the Flexible Side-Chain is "Linked" to the Core:

         \                                                               
          C=O     The Backbone carbonyl group     (CORE)
         /
   --N--CH        The Backbone nitrogen and CA    (CORE)
         \     T1
          CH2     (CB) the beta-carbon group      (LINKER IN CORE)
         /     T2
        CH2       (CG) the gamma-carbon group     (Flexible)
         \     T3
          CH2     (CD) the delta-carbon group     (Flexible)
         /     T4
       HN         (NE) the epsilon-nitrogen group (BEAD)
         \     T5
          C       (CZ) the zita-carbon atom       (BEAD)
         / \
       NH2  NH2   (NH1 & NH2) its amino groups    (both in the BEAD)

In this example there are five rotatable torsions labelled T1 to T5. Another arginine, more deeply buried in the protein, might also have CG assigned by GRIN to the Core. Atom CG would then be the Linker and there would only be four rotatable torsion angles (T2, T3, T4 and T5). Note that the bond to the Linker is regarded as a rotatable torsion.

The assignment of each Target atom as being in the flexible part of the Target, or being in a Bead or the Core, is made automatically by GRIN, according to the local structure of the Target. However, the User can force an atom one way or the other if he or she wants to, so long as he does this carefully (See FORCING ATOMS INTO THE CORE). The assignments work like this:

• THE CORE is the biggest united part of the Target in which conformational flexibility will be neglected. In haem for example, the porphyrin ring is the Core. A Target can only have one Core which is treated as a single immovable object. Target atoms in the Core are identified in the GRINKOUT file by the value NRES = 1

• A BEAD is any other part of the Target in which changes of conformation will be neglected. The NH.C(NH2)2 group at the end of the arginine side-chain is a Bead, because any movement of its heavy atoms RELATIVE TO EACH OTHER is neglected by Grid. However this Bead can change its position and orientation relative to the Core, as the torsion angles alter in the chain of methylene groups which link it to the Core. A Bead can move within limits which are strictly defined by the length of its methylene chain. A Target can have upto 999 Beads in this Version of Grid, and each Bead can have upto 49 atoms. Target atoms in a Bead are identified in the GRINKOUT file by the value NRES = n where n > 1

• The Flexible parts of the Target are outside the Core and outside the Beads. Atoms in flexible parts of the Target are identified in the GRINKOUT file by the value NRES = 0

Of course, this treatment of a Target as "Core+Beads+Chains" is a greatly simplified model, and the algorithms in this Version of the Programmes are still at an embryonic stage of development. They should be used with care, and results obtained when MOVE>0 should be interpreted with particular caution.

The maximum size and number of beads

Programme Grin can handle upto 1000 Beads and upto 50 atoms in each Bead.

Pre-preparation of the Target for Grin

If a flexible side-chain is folded up, so that all its atoms make good Van der Waals contacts with each other and/or with the Core, then it may be treated as part of the Core. If you want the side-chain to be treated as being flexible, we recommend that you extend it away from the Core and away from other side-chains, before using your PDB file as input for GRIN.

21.4.3. The Core

As explained above, every Target must have a Core which is chosen by Programme GRIN. The position of each atom in the Core is fixed, and so the whole Core will be a rigid moiety at a fixed place with a fixed orientation.

GRIN will treat the whole system as the Core by default, if the Target is so flexible that no explicit Core atoms can be identified by GRIN. In this case Target flexibility will be disregarded altogether, unless the User has selected certain atoms as constituting a Core, and has assigned others as being flexible. (See also FORCING ATOMS INTO THE CORE).

21.4.4. Beads

A Bead is also treated as a rigid moiety without conformational flexibility. However each Bead can move and alter its orientation in response to the Probe. It must be linked by a chain of bonds to the Core, but the flexible atoms of the chain move independently whereas the atoms of the Bead all move together as a piece. A Target can contain upto 1000 Beads, each of which may have upto 50 atoms. In normal use the most common beads are the carboxy groups and guanidinium and imadazole groups of proteins.

The methylene groups of Beads are treated as extended atoms; i.e. as spherical units each of which is rather bigger than a carbon atom by itself. For instance one might have a carboxy group embedded in a methylene chain like this:

         __
        /  \__
        \__/  \_CH2.CH2.CH2.CH2.CH2.COO.CH2.CH2.CH2.CH2.CH3
           \__/                     A    m

Bead A is really: CH2.COO but the CH2 is replaced by one spherical extended atom in the bead structure, and the oxygens are treated differently. One is an unionized sp² oxygen O and the other is an sp³ ester oxygen OES. Note that the methylene group m is not part of Bead A because the torsion angle CH2-C-OES-CH2 may not be rigidly fixed.

21.4.5. Flexible atoms and Beads

Flexible atoms and Beads will be treated like this by the Grid Force Field when MOVE=1 in Programme GRIN and MOVE>0 in GRID:

1. GRID starts by dealing with each Bead individually. It uses the Bead as a separate Target, and does a miniature Grid run in order to find how the chosen Probe would best interact with that particular Bead. The main Grid run starts after all the Beads have been studied in this way.

2. During the main Grid run the Bead is always oriented so that it makes its best interaction with the Probe if it can. Thus the most favoured atom of the Bead is turned towards the Probe, and the other atoms are arranged at increasing distances. This is done because of geometrical constraints imposed by the rigid structure of the Bead. For example the best interaction between a guanidinium Bead and a carbonyl oxygen Probe might be represented like this:

   CH2---CH2 H \ N--H \ / N---C O H \ (The Probe) N--H H

   Here the probe is fixed at its grid point, and the Bead has turned so that both NH2 groups are well oriented. However, the third nitrogen of the Bead is inevitably placed less favourably in relation to the Probe, and an allowance must be made for this. Users should note that the algorithms for flexible atoms and Beads have only been tested for a relatively short time, in comparison with several years for the traditional Grid algorithms. The new algorithms may not perform well under all conditions. Please inform Molecular Discovery if you detect any anomalies.

   NB: The concept of a Core was introduced in Version 15 of the Programmes, and Beads were introduced in Version 16. We recommend that Version 22of GRIN should be used with Version 22 of GRUB and Version 22 of GRID. Versions should not be mixed.

3. With the Probe at its grid point, Programme Grid deals with Core atoms in the usual way, using the traditional Grid algorithms for atoms at fixed positions.

4. When Grid comes to a flexible atom or Bead, Grid first attempts to decide if the Probe will be attracted or repelled.

5. If it will be repelled, GRID attempts to move the flexible atom or Bead as far away from the the Probe as possible. It does this while the Probe is located at its predetermined Grid Point. It takes account of the fact that the flexible Target atom or Bead cannot move too far, because it is bonded through a chain of bonds to a Linker Atom in the Core.

6. On the other hand, if the Target atom or Bead will be attracted to the Probe, GRID attempts to move it as close to the the Probe as possible. It does this while the Probe is located at its predetermined grid point. It takes account of the fact that the Target atom or Bead cannot move too far, because it is bonded through a chain of bonds to a Linker Atom in the Core. It tries to move the Target atom until it is at its most favourable distance from the Probe.

7. GRID then computes the interaction energy between that particular flexible Target atom and the Probe, with the Target atom at the appropriate adjusted distance from the Probe. If the Target atom is in a bead, it makes allowance as described above, for the fact that the atom may not be able to get to its most favourable position because of the rigid constraints imposed by the Bead structure.

8. An entropic contribution is then calculated on the basis of three separate factors:

   • How long is the flexible chain of atoms from the interacting Target atom to its Linker.

   • How far is the Probe from the Linker.

   • Is the Flexible chain of atoms attracted to the Probe or repelled.

   For example, if the Probe is attracted to the Target atom, and pulls the side-chain into an all-trans orientation, this is entropically unfavourable and the size of the entropic penalty can be determined because it depends on the number of rotatable torsions. No entropic term is used if the side-chain is repelled from the Probe (although one might argue that the side-chain would again be stretched all-trans).

21.4.6. Parameters for the conformationally flexible model

Information about the atoms which can move will be recorded by this Version of GRIN as the following items in the GRINKOUT record for the atom. These records are then input to GRID, where they are used by the new GRID algorithms:

21.4.7. Different "KINDS" of atoms

Record 1445 is a Special Record in the GRINKOUT file prepared from PDB.PDB. It relates to atom 15 which is CG1 of valine. This CG1 is described as a "KIND 10" atom. The concept of atom "KIND" will now be considered. There are some KINDS of atom in this Version of the Programmes: KIND 10, 11, 12, and 13.

The "TYPE" of an atom has aleady been dealt with in detail. It usually defines the influence of hybridisation upon the arrangement of the atom's hydrogen bonds. The "KIND" of atom defines the way in which the atom is joined by one or more bonds to the Core of the Target. It therefore describes some special property of an atom in a flexible part of the Target.

21.4.8. Special records for "KIND 10" atoms

A KIND 10 atom is joined to the Core of the Target, so that it can only move by torsional rotation in a circular orbit. The associated Special Records are written by FORMAT Statement 140 like ordinary atom records in a GRINKOUT file.

The same variable names will be used in describing the Format of Special Records, but of course the meanings attached to each variable are different in a Special Record. These special meanings depend on the KIND of Special Record, and will now be described for KIND 10 records using Special Record 1445 in the GRINKOUT file from PDB.PDB as an example:

DIFFERENT KINDS OF SPECIAL RECORDS

We are considering a KIND 10 atom, but there are always two different Special Records associated with such an atom. This is because a lot of information is required in order to describe the mobility of a KIND 10 atom.

The first Special Record associated with a KIND 10 atom is known (not surprisingly) as a KIND 10 Special Record. It is always followed immediately by a KIND 11 Special Record containing more information. The variables in a KIND 10 and a KIND 11 Special Record have different meanings which are described in detail below.

KIND 10 SPECIAL RECORDS FOR KIND 10 ATOMS

A KIND 10 atom is constrained by its bonds, so that it may only move in a circular orbit. It is normally linked through one atom to the Core of the Target, and the axis of rotation is defined by the Linker atom and the Core atom to which that Linker is bonded. For example with valine we have:

          \
           C=O     The Backbone carbonyl group   (CORE)
          /
    --N--CA        The Backbone nitrogen and CA  (CORE)
          \
           CB      The beta-carbon atom          (LINKER IN CORE)
          /  \
        CG1   CG2  The gamma-carbon atoms        (KIND 10 ATOMS)

In this example CG1 and CG2 are KIND 10 atoms which can rotate around the CA - CB axis. We suggest that you now generate a GRINKOUT file from file PDB.PDB, if you have not already got one. This PDB.PDB file is available in the tutorials directory, and its GRINKOUT file will be used as an example in the following descriptions. Make sure that directive MOVE = 1 in the command file for GRIN when you prepare GRINKOUT.DAT.

Any atom which has an associated Special Record, will always be marked in the main body of the GRINKOUT file by a plus sign (+) in the last column. Its N6 pointer (the number immediately before the plus sign) points to the Special Record. For example, in the GRINKOUT file from PDB.PDB, there is a plus sign at the end of the record for atom 15, and N6(15) = 1430.

The value N6(15) = 1430 means that the Special Record associated with atom 15 comes 1430 lines further down in the file; i.e. it is record 15 + 1430 = 1445. This layout ensures that the Special Records will all be grouped together near the end of the GRINKOUT file. They are always marked by a plus sign in column 6, and the first two special records in the GRINKOUT file from PDB.PDB are like this:

    1445+  15     1   4.272  18.817  16.652  0.59  1.41 ... 10 14 ...
    1446+  15     0   0.173   0.450   0.876  0.00  0.00 ... 11  0 ...

KIND 10 SPECIAL RECORDS (e.g. Record 1445+ above)

The variables in the KIND 10 Special Record for a KIND 10 atom are interpreted as follows. Note that the Special Records begin immediately after the last hydrogen-bonding hydrogen record.

There are four more integers after N1(I) on the line for a Special Record. Only one of these is used for a KIND 10 Special Record, and this is:

None of the other variables are used in the KIND 10 Special Record of a KIND 10 atom, and all the other numbers are set to zero.

21.4.9. Forcing atoms into the Core

The Target is divided by Programme Grin into a "Core Region" and "Flexible" Side Chains. However, the User may want to force some of the Flexible Side Chain atoms into the Core, and can edit the original PDB input file in order to do this. The PDB file may also be edited in order to get the opposite effect, and force Core atoms to be treated flexibly. Great care is needed when atoms are being selected for forcing, because an inappropriate choice of atoms or incorrect editing may lead to quite unexpected results!!

The occupancy and thermal record fields (OCCUP(I) and BVAL(I)) of a PDB file should have positive or zero values. These values are normally ignored by Programme Grin which uses the same positions for extra data in the GRINKOUT.DAT file as explained above. However, if the OCCUP(I) of an atom is set to -99.99 in the PDB file, then GRIN will try to force that atom into the Core of the Target. On the other hand, if BVAL(I) is -99.99 then Grin will try to force the atom out of the core. These columns should therefore be edited if need be, and the explicit value -99.99 included as a flag where appropriate in order to do the forcing.

Any negative value for OCCUP or BVAL could have been used with previous versions of Programme Grin, in order to force atoms into or out of the Core. However certain software packages create PDB files in which some atoms have been given negative numbers for BVAL or occupancy. The previous versions of Grin would have tried to force such atoms into or out of the Core, which was not what the User intended. This is much less likely to happen with the new Grin Version, since atoms will only be forced when the explicit value -99.99 is specified for OCCUP or BVAL.

Particular care should be taken about the following points:

• OCCUP and BVAL records are not obligatory in a PDB file, and you can just leave the fields empty. But if you do have OCCUP and/or BVAL records, then they must be in the correct columns as specified by the Brookhaven PDB format. The first example shows an ordinary ATOM record with OCCUP = 1.00 and BVAL = 76.48 as observed by the X-ray crystallographer:

ATOM     18  N  SER  14     0.516  10.603  28.970  1.00 76.48

and the next example shows a HETATM record with OCCUP = -99.99 which indicates that the User wants this atom to be forced into the Core:

HETATM  217  N  TRP  28    20.330  18.481  24.976-99.99  0.00 ooo.oobbb.bb

The six character positions assigned for OCCUP, and the six for BVAL, are marked by ooo.oo and bbb.bb on the line which follows the HETATM record.

• Of course one must not have OCCUP = -99.99 and BVAL = -99.99 on the same line!

• The Programme may not be able to force an atom into the Core when you ask it to. This happens most often if you try to force an inappropriate atom into the Core. Therefore one should always check the GRINKOUT.DAT file to be sure what has actually happened.

• If you try to force an extra atom into the Core you may get more than you bargained for, because several other atoms may go into the Core as well.

• In other circumstances, if you try to force an extra atom into the Core, you may not get as many Core atoms as you expected because the Core may be forced into a new and different part of the Target altogether. This happens most often if you try to force an inappropriate atom into the Core. Suppose, for instance, that you started with a big Core on the "right" of the Target when you did no forcing. You might then try to force an extra atom at the "left" of the molecule into the Core as well. At this point you would have a big Core on the right in the natural Core of the Target, and a small forced one-atom Core in an inappropriate place on the left. In this situation, if Grin could not find a way to link these two parts of the Core together, it might have to choose between them. It might then choose the smaller piece of Core on the left, because you were Forcing an atom into that part of the Core. On balance, therefore, you might get the piece of Core which you were forcing, but could loose a lot more of your original Core.

• The Programme may not be able to force an atom out of the Core when you ask it to. (An atom may be left in the Core when you try to force it out, because the necessary algorithms are not yet available for dealing with the flexibility of that particular kind of atom).

In conclusion, therefore, one should always check the GRINKOUT.DAT file particularly carefully if atoms have been forced into the Core or out of it.

21.4.10. Conformationally flexible small molecules

Haem provides an example of a small molecule with a Core region and a pair of side-chains whose conformational movements can be assessed by setting MOVE=1 in the GRIN and GRID runs. Subject to certain reasonable limitations, one should get the same Grid Map from haem when MOVE=1, irrespective of the initial propionate side chain conformations. This is because the side chains will be moved into their appropriate conformations throughout the GRID run (if MOVE=1) without regard to their starting positions.

This means that one may NOT have to select an arbitrary conformation for each molecule, before preparing Grid maps as input data for COMFA-type studies. One could for instance use a set of haem-type molecules, and make the Grid maps for a set of Probes, without defining the exact conformation of each propionate side-chain. Of course it might be better if one knew the conformation of each molecule at its receptor site, but the option MOVE=1 may provide a useful alternative approach when those actual receptor conformations are not known. Further information is provided below, together with the instructions for Directive MOVE in Programme GRID.

Pre-preparation of the Target for Grin

If a flexible side-chain is folded up, so that all its atoms make good Van der Waals contacts with each other and/or with the Core, then it may be treated as part of the Core.  If you want the side-chain to be treated as being flexible, we recommend that you extend it away from the Core and away from other side-chains, before using your PDB file as input for GRIN.

21.4.11. Conformationally flexible Hydrophobicity

Conformationally flexible parts of the Target receive special treatment when directive MOVE has been set in Programme GRID, and the Hydrophobic Probe is being used. This Hydrophobic Probe (which we will call HP) represents a hydrophobic molecule approaching the surface of the Target.

HP might be approaching the hydrophobic surface of the Target at a place where there was a flexible hydrophobic side-chain. In this case it is assumed that good hydrophobic contacts would be made (if possible), and that the side-chain would tend to wrap round HP thus tending to optimise those contacts.

Alternatively HP might be approaching the surface in a generally hydrophobic region, but where there was one polar flexible side-chain. In this different situation it is again assumed that good hydrophobic contacts will be made between Target and Hydrophobic Probe (if possible), but those contacts will now be facilitated if the polar side-chain moves out of the way as far as it can. Thus it may be seen that different types of side-chain are assumed to show different responses to the Hydrophobic Probe, but in every case the response tends to favour the hydrophobic interaction.

21.4.12. Movement of counter-ions

It is sometime necessary to include counter-ions in the PDB file, in order to balance the overall electrostatic charge of the Target. These counter-ions may either be in fixed positions (as defined by their xyz coordinates), or they may be allowed to move in response to the Probe. For further information see above under the headings METAL CATIONS OF THE TARGET and THE POSITION OF DNA counter-ions.