number of peptide fragments. In due time, the studies led to employing all-atom
force fields. Further, most of the force fields are extended to their all-atom versions
with reference to the work done by Weiner [70]. The advent of high-performance
computers has prompted the development of new form of force field which is
popularly known as ff94 force field [71]. The algorithms and automation for force
field beyond protein molecule were achieved by the introduction of antechamber
program, which completely automates the creation of AMBER-like force field for
molecule. The use of fitted charges at the HF/6-31G level has shown a general way to
develop charges for all 20 amino acids in the way which is roughly consistent with
the water molecule. The implementation of the above method for the development of
charges has two major complications: one being the underdetermination of effective
charge in more buried atoms and the other being the procedure of implementing
Restrained Electrostatic Potential Fit (RESP) where the charges depend on the
molecular conformations. To overcome these problems, a more complex term is
required, and this is overcome in ff94 by fitting the charges simultaneously to several
conformations, hoping to achieve optimal average behavior. Torsion angle parameters for the / and w backbone angles affect largely the energies of helices and sheet
in proteins. This was done by the ff94 by fitting representative points on the
dipeptide maps for glycine and alanine and computed at MP2 level with TZP basic
set. In recent years, work is still actively undertaken to test the potentials with the
experimental values of short peptides. From the works of Damm, Mitsutake and
Garcia [72–74], there are modifications done to the ff94, and at least two modifications have been proposed based on large-scale short peptide simulations [75, 76].
CHARMM force field
Chemistry at Harvard using Molecular Mechanics (CHARMM) program [77] was
developed in 1983. There were rather many versions of CHARMM, namely
CHARMM19, CHARMM22, and CHARMM27, with different charge-deriving
methodologies incorporated. The same difficulty of torsion angle parameterization
exits in CHARMM as observed in case of ff94. The comparison of AMBER and
CHARMM force fields in respect to protein potential shows that the peptide carbonyl group is less polar in CHARMM22 than with ff94, and NH dipole is less
polar in the case of AMBER. We can observe the similarities in the charge model
between the two force fields prominently than the differences between them.
OPLS force field
This is the other force field developed during the 1980s by Jorgensen and co-workers
to simulate liquid state. This force field called optimized potential for liquid simulations (OPLS) plays a great importance to non-bonded interactions by comparison
with liquid-state thermodynamics [69]. For proteins, polar hydrogen model only was
developed initially with the atoms type and valance (bond, angle, dihedral) parameters form AMBER, and later, all-atoms force field was developed (OPLS-AA).
Importance of protein flexibility
SBDD works by the identification of sites on the protein surfaces, named “hot
spots,” where the chemical motifs will bind the receptor residues. The methods
Structure-Based Drug Design…
281
force fields. Further, most of the force fields are extended to their all-atom versions
with reference to the work done by Weiner [70]. The advent of high-performance
computers has prompted the development of new form of force field which is
popularly known as ff94 force field [71]. The algorithms and automation for force
field beyond protein molecule were achieved by the introduction of antechamber
program, which completely automates the creation of AMBER-like force field for
molecule. The use of fitted charges at the HF/6-31G level has shown a general way to
develop charges for all 20 amino acids in the way which is roughly consistent with
the water molecule. The implementation of the above method for the development of
charges has two major complications: one being the underdetermination of effective
charge in more buried atoms and the other being the procedure of implementing
Restrained Electrostatic Potential Fit (RESP) where the charges depend on the
molecular conformations. To overcome these problems, a more complex term is
required, and this is overcome in ff94 by fitting the charges simultaneously to several
conformations, hoping to achieve optimal average behavior. Torsion angle parameters for the / and w backbone angles affect largely the energies of helices and sheet
in proteins. This was done by the ff94 by fitting representative points on the
dipeptide maps for glycine and alanine and computed at MP2 level with TZP basic
set. In recent years, work is still actively undertaken to test the potentials with the
experimental values of short peptides. From the works of Damm, Mitsutake and
Garcia [72–74], there are modifications done to the ff94, and at least two modifications have been proposed based on large-scale short peptide simulations [75, 76].
CHARMM force field
Chemistry at Harvard using Molecular Mechanics (CHARMM) program [77] was
developed in 1983. There were rather many versions of CHARMM, namely
CHARMM19, CHARMM22, and CHARMM27, with different charge-deriving
methodologies incorporated. The same difficulty of torsion angle parameterization
exits in CHARMM as observed in case of ff94. The comparison of AMBER and
CHARMM force fields in respect to protein potential shows that the peptide carbonyl group is less polar in CHARMM22 than with ff94, and NH dipole is less
polar in the case of AMBER. We can observe the similarities in the charge model
between the two force fields prominently than the differences between them.
OPLS force field
This is the other force field developed during the 1980s by Jorgensen and co-workers
to simulate liquid state. This force field called optimized potential for liquid simulations (OPLS) plays a great importance to non-bonded interactions by comparison
with liquid-state thermodynamics [69]. For proteins, polar hydrogen model only was
developed initially with the atoms type and valance (bond, angle, dihedral) parameters form AMBER, and later, all-atoms force field was developed (OPLS-AA).
Importance of protein flexibility
SBDD works by the identification of sites on the protein surfaces, named “hot
spots,” where the chemical motifs will bind the receptor residues. The methods
Structure-Based Drug Design…
281