A comprehensive review of computational techniques for the prediction of drug side effects

Abstract

Drugs refer to the chemical compounds that are consumed by the human body and induce a change by interacting with the protein targets. The drugs may induce favorable or unfavorable changes in the human body. The unfavorable changes that are elicited by the drugs in the human body are known as drug side effects. These side effects range from minor reactions like headache to serious reactions such as cardiac arrest, cancer, or even death. The drugs are tested for their side effects based on laboratory experiments. However, these experiments are costly as well as prolonged. An alternative to the laboratory experiments is provided through the computational methods. Many computational techniques for identifying the drug side effects have been developed in the recent past. This review aims to summarize the important studies and contributions in the field of drug side effect prediction by computational techniques. A description of the data sets related to drug side effects and the metrics used for the evaluation of drug side effect prediction methods have also been explained. This article also highlights the future research that can be undertaken in this field. To the best of our knowledge, this is the first extensive review of the computational methods that have been developed for drug side effect prediction.

Molecular Docking: Metamorphosis in Drug Discovery

“…When a medicine attaches to a certain protein, it can produce side effects. Drug side effect prediction using docking-based techniques Classification of drug side effect prediction approaches [143].…”

Clustering Protein Binding Pockets and Identifying Potential Drug Interactions: A Novel Ligand-based Featurization Method

“…41 Recent advances in the toxicity prediction space also include neural fingerprinting, 42 conformal prediction, 43 and several others. [44][45][46] While our approach doesn't delve into predicting ADRs, we expect that our simple, granular featurization approach will be useful in multi-target and toxicity prediction.…”

Comparative Proteomics and Genome-Wide Druggability Analyses Prioritized Promising Therapeutic Targets against Drug-Resistant Leishmania tropica

“…Multiomics data, such as proteomics, metabolomics, and genomics, have proven tremendous in drug discovery processes by significantly lowering the cost and time required for in vivo and experimental screening [ 12 ]. Subtractive and comparative genomic approaches screen the entire proteome of the host and pathogen to prioritize pathogen-specific essential proteins holding therapeutic potential [ 13 , 14 ]. In the present study, we analyzed the recently updated L. tropica genome resources and prioritized several druggable protein targets.…”