refmac5 XYZIN foo_cycle_i.brk HKLIN foo.mtz
TLSIN tlsin.txt TLSOUT tlsout.txt
XYZOUT foo_cycle_j.brk HKLOUT foo_cycle_j.mtz
REFMAC now has two 'modes': REVIEW which only checks and updates the input model to establish that the geometric restraints can be properly set up, and the standard REFINE mode documented in keywords.
This replaces and extends one of the roles of PROTIN. The input model (or PDB file) will contain coordinates identified by a CHAIN identifier, a RESIDUE (or more precisely, a monomer) identifier (e.g. ALA TYR HEM HOH) and an atom name (e.g. CA or Fe+3). The output file will have the following extra information.
REFMAC now does a preliminary distance angles check on the input coordinates and suggests likely di-sulphide bonds and other links. These are written to the output PDB file in the standard PDB format:
SSBOND 1 CYS A 48 CYS D 48 1555 2555 SSBOND 2 CYS B 48 CYS C 48 1555 2555 LINK 1 OE1 GLU A 276 CA CA W 1 1555 1555 LINK 1 OE1 GLU A 276 CA CA W 1 1555 1555 LINK 1 OD2 ASP A 307 CA CA W 1 1555 2556 LINK 1 O HOH W 133 CA CA W 1 1555 1555
The LINK record can be modified by the user to include the target distance. In this format: the distance has format F10.5.
LINK 1 OE1 GLU A 276 2.13400 CA CA W 1 1555 1555 LINK 1 OE1 GLU A 276 2.13400 CA CA W 1 1555 1555 LINK 1 OD2 ASP A 307 2.13400 CA CA W 1 1555 2556 LINK 1 O HOH W 133 2.40000 CA CA W 1 1555 1555
REFMAC now does a preliminary survey of cis-peptides and writes the records to the output PDB file.
CISPEP 1 SER A 26 PRO A 27 0 0.00 CISPEP 1 HIS A 220 PRO A 221 0 0.00
REFMAC does not output the SEQRES records to the XYZOUT file. All REMARK records are stripped out; those relevant to refinement are replaced.
All this output needs to be carefully checked, and edited by the user.
If this is enough to describe the chemistry of the model completely, the REFINE mode of refmac can be started without any other action. However, if the molecule under study has a novel group REFMAC in REVIEW mode will exit with a message requiring the user to provide a complete chemical description of this.
Obviously, before starting restrained refinement it is necessary that the chemical properties of each monomer in the model (i.e. the identifier and charge for each atom, the connectivity and bond types with target distances and standard uncertainities, the angles, chiral centres, planar groups, target torsion angles) are described. For the standard peptides, sugars, and nucleotides plus some common ligands these are provided in distributed library files in $CLIBD_MON/. For any other monomers the user must provide their own library.
The REFMAC program can carry out rigid body, tls, restrained or unrestrained refinement against Xray data, or idealisation of a macromolecular structure. It minimises the coordinate parameters to satisfy either a Maximum Likelihood or Least Squares residual. There are options to use different minimization methods. (At the moment only CGMAT is active.) REFMAC also produces an MTZ output file containing weighted coefficients for SigmaA weighted mFo-DFcalc and 2mFo-DFcalc maps, where "missing data" have been restored.
[Program organisation and communication with it is illustrated in the EXAMPLES section.]