Keywords Computational drug discovery Á Free energy of binding
Hybrid QM/MM Á QM fragmentation Á Binding affinity Á Pharmacokinetic
(PK) properties Á Machine learning approach
Abbreviations
FMO
Fragment molecular orbital
MAO-B
Monoamine oxidase B
MM-GBSA Molecular mechanics–Generalized Born Surface Area
MM-PBSA Molecular mechanics–Poisson–Boltzmann Surface Area
PD
Pharmacodynamic
PK
Pharmacokinetic
QM/MM
Quantum mechanics/molecular mechanics
1 Introduction: Drugs and Targets
Disease can be defined as an abnormal condition that alters the function or behaviour of an organism and this can be caused by different factors, i.e. internally
e.g. due to the presence of disease-causing genes or due to external factors.
Externally, disease may be caused due to malnutrition or subjecting a human to
severe external conditions such as exposing to radiation or pollution or microbial
infections or severe physiological conditions which leads to damage or malfunctioning of body machineries. Thanks to genomic analysis of normal and diseased
persons, we know that the protein profile appears quite different in these two cases
and by targeting the biomacromolecules expressed in the diseased state, and we can
develop methods to arrest the progress of the disease. By comparative protein
profiling of normal and diseased persons or by comparing the genomes of human
and pathogenic micro-organisms, [1–3] we already know the information about the
potential targets, but then the problem lies in identifying whether the aberrant
expression of a certain biomacromolecule is the main cause of disease or may be a
side product of another key process. Once the key target (protein or enzyme) is
identified, primay task is to design small molecules that can modulate the target (this can be either of inhibitor, substrate, inducer). Subsequently, the active
compound (also called hit molecule) is further optimized to pre-clinical candidate.
The aim of lead optimization is not only to increase the potency, but also to
reduce any off-target binding. In this chapter, we will discuss how to use computational approaches not only to identify small molecules that can inhibit or modulate the catalytic process of a key enzyme that is connected with disease, but also
to understand the fundamental process of biomolecular recognition which assists in
the lead optimization process in the drug discovery and development projects. The
properties of the ligand to be optimized are binding affinity and specificity towards
a key target biomacromolecule. These target molecules can be located within
222
N. A. Murugan et al.
Hybrid QM/MM Á QM fragmentation Á Binding affinity Á Pharmacokinetic
(PK) properties Á Machine learning approach
Abbreviations
FMO
Fragment molecular orbital
MAO-B
Monoamine oxidase B
MM-GBSA Molecular mechanics–Generalized Born Surface Area
MM-PBSA Molecular mechanics–Poisson–Boltzmann Surface Area
PD
Pharmacodynamic
PK
Pharmacokinetic
QM/MM
Quantum mechanics/molecular mechanics
1 Introduction: Drugs and Targets
Disease can be defined as an abnormal condition that alters the function or behaviour of an organism and this can be caused by different factors, i.e. internally
e.g. due to the presence of disease-causing genes or due to external factors.
Externally, disease may be caused due to malnutrition or subjecting a human to
severe external conditions such as exposing to radiation or pollution or microbial
infections or severe physiological conditions which leads to damage or malfunctioning of body machineries. Thanks to genomic analysis of normal and diseased
persons, we know that the protein profile appears quite different in these two cases
and by targeting the biomacromolecules expressed in the diseased state, and we can
develop methods to arrest the progress of the disease. By comparative protein
profiling of normal and diseased persons or by comparing the genomes of human
and pathogenic micro-organisms, [1–3] we already know the information about the
potential targets, but then the problem lies in identifying whether the aberrant
expression of a certain biomacromolecule is the main cause of disease or may be a
side product of another key process. Once the key target (protein or enzyme) is
identified, primay task is to design small molecules that can modulate the target (this can be either of inhibitor, substrate, inducer). Subsequently, the active
compound (also called hit molecule) is further optimized to pre-clinical candidate.
The aim of lead optimization is not only to increase the potency, but also to
reduce any off-target binding. In this chapter, we will discuss how to use computational approaches not only to identify small molecules that can inhibit or modulate the catalytic process of a key enzyme that is connected with disease, but also
to understand the fundamental process of biomolecular recognition which assists in
the lead optimization process in the drug discovery and development projects. The
properties of the ligand to be optimized are binding affinity and specificity towards
a key target biomacromolecule. These target molecules can be located within
222
N. A. Murugan et al.