from marine snail [156, 157] were identified. The advent of proteomic data analysis
with Mass Spectroscopy (MS) and Next-Generation Sequencing (NGS) helps in
identification of such potential peptides which are conserved evolutionary. Proper
screening of such peptides can be helpful in the identification of peptides which
might be closer to lower end of molecular weight spectrum (500 Da) and far from
the higher end (5000 Da) and understanding structure–function relationship of
bioactive sequences. Such systematic search can yield side effect-free peptide or
peptide-like molecules with potential pharmaceutical application.
Other approaches in exploration of potential lead molecule employing different
approaches and revisiting old ones with new perspective will help in SBDD, and
one such aspect is drug repurposing. Drug repurposing or drug repositioning is a
process in which new uses are identified and attributed for a drug which already
exists [158]. With a huge set of drug molecules in the clinical and preclinical stages,
trials will not be carried out for next level because of side effects or toxicity, etc.
Researchers have started to understand the toxic effect of the drug-like molecule
and maintain the database with these prediction results [159]. The concept of drug
repurposing was introduced to understand the properties of existing drugs for being
safe [160]. Drug repurposing helps in avoiding adverse effects and also in saving
time in drug development pipeline [161], and this is achieved by analyzing the
similarities of drugs. Such similarities provide with a situation, where they inhibit
the same target and result in the same function. The profiling of pharmacological
properties of drugs which are presently available was done by Cheng et al., by
comparing the chemical similarity of drug molecules and their phenotypic effects
[162]. The larger information available deals not only with the properties of drugs
but also regarding targets, and this has resulted in “big data” analysis, network
analysis, machine learning by Bayesian statistics [163], deep learning, multi-task
learning [164], and ligand-based chemogenomic analysis [163, 165, 166], which
are advanced and recent algorithms, and routines are employed in drug designing
approaches. Yet, a lot of scopes still persist and also large information is to be
explored for a better profiling and understanding of the drugs and their physiological roles. Out of many aspects, drug–drug interaction (DDI) is studied in recent
years. This will help in understanding the effect of one drug on the other and their
interaction. Networking analysis among drugs can help in grouping drugs which
might manifest similar physiological effect. Different parameters can be computed
with benchmark parameter with which the dataset can be classified and even be
validated using training and test datasets. Different statistical methods are
employed. For instance, Tanimoto coefficient [167] was employed in understanding
the DDI among 6711 drugs collected from the DrugBank [168]. Similarity analysis,
association networking for scoring the drugs, and target prediction to predict the
numbers of targets each drug binds and target prediction with chemical structure of
drug molecules were performed. This approach will provide us with parameters,
based on which two drugs interact, and the results project that the DDI effect can
manifest due to pharmacokinetic parameters. Even though DDI can help in predicting targets, proposing new targets, and providing us with drug–drug synergetic
effect on same target, there are still many hurdles posed for employing this method.
Structure-Based Drug Design…
289
with Mass Spectroscopy (MS) and Next-Generation Sequencing (NGS) helps in
identification of such potential peptides which are conserved evolutionary. Proper
screening of such peptides can be helpful in the identification of peptides which
might be closer to lower end of molecular weight spectrum (500 Da) and far from
the higher end (5000 Da) and understanding structure–function relationship of
bioactive sequences. Such systematic search can yield side effect-free peptide or
peptide-like molecules with potential pharmaceutical application.
Other approaches in exploration of potential lead molecule employing different
approaches and revisiting old ones with new perspective will help in SBDD, and
one such aspect is drug repurposing. Drug repurposing or drug repositioning is a
process in which new uses are identified and attributed for a drug which already
exists [158]. With a huge set of drug molecules in the clinical and preclinical stages,
trials will not be carried out for next level because of side effects or toxicity, etc.
Researchers have started to understand the toxic effect of the drug-like molecule
and maintain the database with these prediction results [159]. The concept of drug
repurposing was introduced to understand the properties of existing drugs for being
safe [160]. Drug repurposing helps in avoiding adverse effects and also in saving
time in drug development pipeline [161], and this is achieved by analyzing the
similarities of drugs. Such similarities provide with a situation, where they inhibit
the same target and result in the same function. The profiling of pharmacological
properties of drugs which are presently available was done by Cheng et al., by
comparing the chemical similarity of drug molecules and their phenotypic effects
[162]. The larger information available deals not only with the properties of drugs
but also regarding targets, and this has resulted in “big data” analysis, network
analysis, machine learning by Bayesian statistics [163], deep learning, multi-task
learning [164], and ligand-based chemogenomic analysis [163, 165, 166], which
are advanced and recent algorithms, and routines are employed in drug designing
approaches. Yet, a lot of scopes still persist and also large information is to be
explored for a better profiling and understanding of the drugs and their physiological roles. Out of many aspects, drug–drug interaction (DDI) is studied in recent
years. This will help in understanding the effect of one drug on the other and their
interaction. Networking analysis among drugs can help in grouping drugs which
might manifest similar physiological effect. Different parameters can be computed
with benchmark parameter with which the dataset can be classified and even be
validated using training and test datasets. Different statistical methods are
employed. For instance, Tanimoto coefficient [167] was employed in understanding
the DDI among 6711 drugs collected from the DrugBank [168]. Similarity analysis,
association networking for scoring the drugs, and target prediction to predict the
numbers of targets each drug binds and target prediction with chemical structure of
drug molecules were performed. This approach will provide us with parameters,
based on which two drugs interact, and the results project that the DDI effect can
manifest due to pharmacokinetic parameters. Even though DDI can help in predicting targets, proposing new targets, and providing us with drug–drug synergetic
effect on same target, there are still many hurdles posed for employing this method.
Structure-Based Drug Design…
289