microorganism or within the host organism depending upon whether the disease
belongs to infectious or autoimmune category disease, respectively.
2 Optimization of Drug-Likeness
In addition to binding affinity and specificity, there are certain other properties
which are to be optimized for an effective drug i.e. low toxic with improved
potency and orally bioavailable for conventional dosage forms. These properties are
absorption (A), distribution (D), metabolism (M), excretion (E) and toxicity (T),
and they collectively are called ADMET or pharmacokinetic (PK) properties. The
properties in general refer to kinetic behaviour of drugs within body and give
information about the timescale required for the drug to reach the potential target
and lifetime within host organism before removal through excretion (this can be
shortly described as “what the body does to a drug?”). The optimization of potency
(binding affinity) and then the subsequent optimization of pharmacokinetic behaviour have been the major contributing factors for the failures at the phase II and
phase III clinical trials [4–6]. So, it is necessary simultaneously to optimize the
potency along with ADMET properties [7]. There are also other properties that are
essential for oral bioavailability such as solubility and transport properties like
membrane permeability (both cellular and across blood-brain barrier).
Overall, it is apparent that drug design is challenging as we need to optimize
several properties at the same time [3, 6]. In certain cases, optimizing one property
may lead to unexpected changes in another property making this optimization
very complex , and in those cases we need to compromise on certain properties and
try to balance different properties for better PD and PK profile. For example, if the
potency of a drug is superior/outstanding but then if it has very poor PK and PD
properties, then one can use suitable drug delivery systems such that the drug is
delivered to its target biomacromolecule. Given that drug design is an optimization
process, it is inevitable to avoid the use of computers as they can be used effectively
to speed up the overall process. But the only requirement is that we need to have
accurately enough methods that can be written in a numerically solvable form and
can reliably describe these processes involved in the drug association with a biological target and its pharmacokinetics [8]. In general, quantum mechanics is the
fundamental theory which can be used to describe any atomic and molecular systems
and their association process and their response to any external variables like heat,
pressure and fields and to any change in physiological conditions such as pH and
ionic strength. However, the complexity of the mathematics involved in solving the
Schrödinger equation grows with powers of n which is number of basis functions
used to describe one electron orbital used to build wave function of the molecule that
describes its energy and all other properties. For example, the computational demand
is at the power of three in the case density functional theory and can go to power of
5–8 for the theories which can treat the electron correlation more accurately. When
the system size is comparably larger than the wavelength of light and when we are
Recent Advancements in Computing Reliable Binding Free Energies …
223
belongs to infectious or autoimmune category disease, respectively.
2 Optimization of Drug-Likeness
In addition to binding affinity and specificity, there are certain other properties
which are to be optimized for an effective drug i.e. low toxic with improved
potency and orally bioavailable for conventional dosage forms. These properties are
absorption (A), distribution (D), metabolism (M), excretion (E) and toxicity (T),
and they collectively are called ADMET or pharmacokinetic (PK) properties. The
properties in general refer to kinetic behaviour of drugs within body and give
information about the timescale required for the drug to reach the potential target
and lifetime within host organism before removal through excretion (this can be
shortly described as “what the body does to a drug?”). The optimization of potency
(binding affinity) and then the subsequent optimization of pharmacokinetic behaviour have been the major contributing factors for the failures at the phase II and
phase III clinical trials [4–6]. So, it is necessary simultaneously to optimize the
potency along with ADMET properties [7]. There are also other properties that are
essential for oral bioavailability such as solubility and transport properties like
membrane permeability (both cellular and across blood-brain barrier).
Overall, it is apparent that drug design is challenging as we need to optimize
several properties at the same time [3, 6]. In certain cases, optimizing one property
may lead to unexpected changes in another property making this optimization
very complex , and in those cases we need to compromise on certain properties and
try to balance different properties for better PD and PK profile. For example, if the
potency of a drug is superior/outstanding but then if it has very poor PK and PD
properties, then one can use suitable drug delivery systems such that the drug is
delivered to its target biomacromolecule. Given that drug design is an optimization
process, it is inevitable to avoid the use of computers as they can be used effectively
to speed up the overall process. But the only requirement is that we need to have
accurately enough methods that can be written in a numerically solvable form and
can reliably describe these processes involved in the drug association with a biological target and its pharmacokinetics [8]. In general, quantum mechanics is the
fundamental theory which can be used to describe any atomic and molecular systems
and their association process and their response to any external variables like heat,
pressure and fields and to any change in physiological conditions such as pH and
ionic strength. However, the complexity of the mathematics involved in solving the
Schrödinger equation grows with powers of n which is number of basis functions
used to describe one electron orbital used to build wave function of the molecule that
describes its energy and all other properties. For example, the computational demand
is at the power of three in the case density functional theory and can go to power of
5–8 for the theories which can treat the electron correlation more accurately. When
the system size is comparably larger than the wavelength of light and when we are
Recent Advancements in Computing Reliable Binding Free Energies …
223