Next: , Up: Degrees of Freedom


12.5.1 Backbone Degree of Freedom

The backbone degree of freedom constructs the backbone atoms of a polypeptide. It can handle all the natural amino acids as well amino-isobutyric acid (AIB). For our purposes, the backbone atoms are N, H, CA, CB, C, and O and the ring atoms in proline.

This degree of freedom is implemented like a macro in that the program treats the backbone of each residue as a separate degree of freedom. Thus, when this command is used to generate conformations for n backbones, n backbone degrees of freedom are generated internally.

The STARTRES option specifies the starting residue for which a backbone is to be constructed. The LASTRES option specifies the final residue for construction. Together with the FORWARD and REVERSE options, these options also specify which way the construction should be done, i.e. constructing from N to C or C to N. The following decision table gives the order used for linear structures:

                            FORWARD       REVERSE       NONE
          STARTRES<LASTRES   N->C          C->N         N->C
          STARTRES=LASTRES   N->C          C->N         N->C
          STARTRES>LASTRES   N->C          C->N         C->N

The following decision table gives the order used for cyclic structures:

                            FORWARD       REVERSE       NONE
          STARTRES<LASTRES   N->C          C->N     Use shortest distance.
          STARTRES=LASTRES   N->C          C->N     Forward = true
          STARTRES>LASTRES   N->C          C->N     Use shortest distance.

The range of residues cannot span between two segments of the PSF.

The MAP option specifies the Fortran unit number for the backbone energy map, see Backbone Maps, to be used. If no map is specified, then the default map unit is used. The defaults are specified by the PROMAP, GLYMAP, and ALAMAP global options, see Backbone and Chain Closure Overview. The EMAX option specifies the maximum energy of entries in the map for the range of residues. If no EMAX option, then the global EMAX option appropriate for the given residue applies. You can use this option to override the default EMAX option for a particular backbone specification.

The MAXEVDW option specifies the maximum energy for any repulsive contact between any generated atom and atoms in the surroundings.

The CISTRANS and ALLCISTRANS options control whether cis peptides are included in the search. By default only trans peptides are used. CISTRANS specifies that only prolines may have cis peptides. ALLCISTRANS specifies that any amino acid may have cis peptides. In this context, CISTRANS and ALLCISTRANS applied to residue n refer to the peptide bond between residue n-1 and n. N.B. If no cis peptides are included in the torsion angle map, then this option has no effect. Generally, all the peptide maps defined over a grid have both cis and trans peptides, but you should check if this option is important to your problem.

The CLSA, CLSD, and MAXDT options control an important optimization in the search. Since this degree of freedom is generally used in conjunction with a chain closure, there is no point constructing backbones that stray too far for closure to take place. The CLSA option specifies the atom which will terminate the chain closure. For a construction in the N->C direction, this atom should be a CA; for the opposite direction, this atom should be an N. The program will construct a model backbone, all torsion angles being trans, and all bond angles involved in chain closures being stretched by MAXDT degrees (see below for more details) which would connect from the residue whose backbone is being searched to that CLSA atom, and it will measure this distance. If a backbone conformation is constructed with a separation distance greater than the CLSA distance, that search path will be abandoned. CLSD permits the closing distance to be modified. If just a number is specified with CLSD, then the program uses this value in place of the model backbone calculation. If the keyword, DELTA, is also used, then the real number is added to the model distance.

The construction of the model peptide attempts to correctly account for the flexibility permitted in peptide bond angles. If MAXDT is specified, then all bond angles in the model calculation are stretched by this amount. If MAXDT is not specified in the backbone command, then CONGEN searches through all the degrees of freedom to see which involve chain closure. If chain closures are found, then the associated bond angles are stretched accordingly. If the search involves no chain closures, then the default value for MAXDT will be used to stretch all bond angles.

The backbone degree of freedom also handles backbones at the N and C termini. The N terminus and C terminus cannot use the standard torsion angle maps because there are fewer atoms than assumed in the construction of the maps and because of the rotational symmetry present at each end. Thus, the CONGEN command will perform a complete sampling for each end using only the MAXEVDW test. The grid for this sampling will be set to value of the GRID option. Also, the rotational symmetry of the terminus will be used to reduce the search unless NOTERSYMMETRY is specified.

N.B., there is currently an error in the design of the code with regard to the construction of prolines at the amino terminus when using the AMBER94 potential. The problem manifests itself in geometric errors in the amino terminal nitrogen. Until this problem is fixed, you should avoid such constructions.

The FIX option is used to specify that the backbone should be constructed in a single fixed conformation without searching. The bond lengths, angles, and torsions for all the atoms in the degree of freedom are calculated from the current coordinates. The normal test of van der Waals overlap applies.