Chemistry Reference
In-Depth Information
could be considered for initial bioisosteric replacements. Another limitation of
the literature search is that the list of possible interchangeable fragments is very
specific to bioisosters that had already been identified. Nevertheless, there is a
need to identify novel possible bioisosteres that had not been identified in order to
generate structural diversity. Performing this task, we can highlight experimental
and computational approaches. During experimental approaches. the synthesis of
diverse analogues of a particular molecular scaffold, limiting the variations for
only one substituent, should be combined with biological evaluations in order to
identify which fragments could be bioisosters of the original lead compound
substituent. The major limitations of this strategy are the cost and time consumed
to synthesize and evaluate an extensive series of bioactive molecules. In this way,
computational approaches are extremely useful tools that can help the search for
novel bioisosters [5, 8, 9] by measurements of shape and electrostatic similarity
among groups of possible interchangeable fragments and guide the synthesis of
novel analogues, reducing time and costs of the process where only the molecules
generated by the most probable successful bioisosteric replacements are
synthesized. It is important to remember that the success of a bioisosteric
replacement is not centered only in maintaining biological activity but the original
parameters of the lead compound should be considered which must be modified,
such as toxicity, metabolic stability, solubility, synthetic accessibility and/or
others.
BIOISOSTERISM AND COMPUTATIONAL APPROACHES
Appropriate bioisosteric replacement is a challenging task that requires a
considerable amount of medicinal chemistry experience. Even if this experience is
available, the identification of a bioisosterically suitable group with an optimal
balance of steric, hydrophobic, electronic and hydrogen bonding properties
usually requires an intensive procedure of trial and error. In the last years a wide
range of diverse and innovative computational approaches are able to assist these
replacements has been developed [5].
In silico
methods can be very useful in the
search for chemical groups sharing similar properties. These methods apply
various chemoinformatic techniques based on bioactivity guided database mining,
defined by the characterization of groups by a range of calculated descriptors and
identification of bioisosteric pairs based on similarity of properties [3].
