Computational Approaches in the Design of Synthetic Receptors - Designing Receptors for the Next Generation of Biosensors - page 138

Table. 1 (continued)
Process Application Targets and references

Neural network using the back
propagation algorithm
(WEKA)

Prediction of imprinting factor
of MIPs and study of
template monomer
complexes

Atropine and Boc- L -Trp
d-Brompheniramine
[ 75 , 76 ]

General chemometric design Optimization for MIP

preparation

Bisphenol-A [ 122 ]

Chemometric multivariate
analysis study on screening
and rebinding using radio-
ligand counting methods

Extensive screening and more
sophisticated response
parameters for analysis

Propranolol [ 123 ]

calculations on multicomponent systems. While molecular modeling of complex
systems and possible interactions of polymers with template, solvent, and other
molecules are difficult because of the requirements of large computational
workloads, the method developed at Cranfield University can achieve this by
simplifying the model. Since the structure of the monomers-template complexes
formed in the monomer mixture is preserved in the synthesized polymer, instead of
modeling the polymer, modeling the monomer mixture and the interactions taking
place in solutions between monomers, cross-linker, template, and solvent would be
possible, which substantially reduces computational load [ 61 - 63 ].
The protocol developed at Cranfield starts with the design of a virtual library of
molecular models of functional monomers and template (Fig. 2 ). The next step is to
screen the virtual library against a template to determine the monomers that
strongly bind to the template. Calculations are performed to estimate how the
monomers bind to template using simulated annealing to determine optimum ratios
of template to monomers. In effect, the strength and type of interactions, existing
between monomers and template in monomer mixtures, which in theory, determine
the recognition properties of the MIP, will be analyzed and used for optimization of
polymer composition. Each of the steps has been described below.

2.1.1 Modeling of the Template Molecule

A molecular model of the template molecule is made and charges for each atom are
calculated, and the structure of the template is refined using MM.

2.1.2 Construction of the Monomer Database

While there are about 4,000 polymerizable compounds that have been reported that
could potentially be used as functional monomers, in reality many of them have
similar properties and functions and hence it is assumed that it is sufficient to
test possible interactions between a minimal library of functional monomers and
a target template [ 124 ]. Several different software packages can be used for creating