Computational Approaches in the Design of Synthetic Receptors - Designing Receptors for the Next Generation of Biosensors - page 145

simulations were carried out with AMBER 7, as it previously showed satisfactory
performance of this force field in the evaluation of the stability of hydrogen bonded
as well as van der Waals adducts [ 143 - 146 ]. The structures of THO, MAA, and
MMA molecules were firstly optimized using DFT [ 147 - 149 ], and their atomic
charges were determined with RESP [ 150 , 151 ]. All the simulations were
performed in the NPT ensemble using Berendsen thermostat and barostat [ 152 ]
with temperature set to 310 K and pressure to 1 atm. THO molecule was surrounded
by functional monomer shells and solvated, creating around it a rectangular paral-
lelepiped acetonitrile box [ 153 , 154 ].
Docking procedures, used to find favorable orientations of the ligands inside the
polymer cavity, were performed using the DOCK5.0 program [ 155 - 157 ], and the
GRID program (GRID, 2004) was used to map the energetic interactions. The created
model was able to predict binding affinity and selectivity when considering THO
analogues, such as caffeine, theobromine, xanthine, and 3-methylxanthine [ 91 ].
The entire modeling study was performed in four different phases: first, a non-
covalent phase where the template and the functional monomers form non-covalent
complexes in solution prior to polymerization; second, a locking phase where the
non-covalent monomers-template complexes are cross-linked and the binding site
is generated with appropriately oriented functional monomers and model polymer
structures are created and selected; third, a validation phase where the polymer
specificity and recognition capabilities are tested; and finally the mapping phase
where the characteristics of the binding cavities are analyzed. This work showed
that the MD simulations were able to predict the selectivity and binding affinity of
the MIP, and when complemented with experimental data gave a clearer picture of
the system and the type of interactions in the complex (Fig. 4 ).
Similarly, selective adsorption properties of dimethoate imprinted polymers
were studied through a MD simulation [ 101 ]. The MD modeling was carried out
to investigate the recognition mechanism by predicting the interaction energy
differences and indicating the active site groups, which confirmed that the MIP,
based on butyl methacrylate (BMA) functional monomer, had the most selective
recognition for dimethoate compared to other functional monomers, including
methyl methacrylate (MMA) and ethyl methacrylate (EMA) (Fig. 5 ).
Two parameters: the imprinting factor indicator and the competitive factor
indicator were calculated, based on the interaction energy differences for the
template in relation to structurally related organophosphorus pesticides (OPs).
The methodology was interesting, in that PRODRG software was used to analyze
polymer topology and MD simulations were performed with the GROMACS-3.2 in
the NPT ensemble with GROMOS 96 force field. A van der Waals cutoff of 1.4 nm
and PME were used to describe nonbonded interactions [ 158 - 161 ]. However, the
influence of the cross-linking agent was neglected, apparently to simplify the
modeling process. A similar experiment was performed by other researchers [ 95 ].
In an effort to obtain the optimized conformation of MIP, the authors saturated the
system of polymer chain and dimethoate with the solvent THF and performing
simulation for 20 ns. Computational prediction was verified through

chromatography.