A Comprehensive Analysis of Artificial Intelligence-Based Drug Side Effect Prediction Techniques

View at publisher Summarise with VIRAT
Authors: Naga Padmaja Yenireddy, G Sri Lakshmi
Cited by: 0
References: 17

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Predicting drug side effects before they reach patients is one of the most important challenges in modern healthcare. Many recent studies have used different artificial intelligence methods to identify possible harmful reactions, but their performance varies widely because they rely on different types of data, learning techniques, and evaluation settings. This work presents a detailed review of these approaches to understand how well they perform, where they succeed, and where they fall short. The analysis covers models based on chemical structures, similarity networks, machine-learning classifiers, deep-learning architectures, and graph-based frameworks. The review shows that while these methods can detect common side effects with reasonable accuracy, they often struggle with rare or complex reactions due to limited data, narrow feature representation, and high computational demands. The study also highlights gaps in coverage across different disorder groups and points out the need for models that combine chemical, biological, and clinical information. This comprehensive assessment helps outline the strengths and weaknesses of current techniques and offers guidance for building more reliable and efficient drug side-effect prediction systems in the future.

Fields of Study & Subjects

Computational Drug Discovery Methods Pharmacovigilance and Adverse Drug Reactions Machine Learning in Healthcare

Publication details

Journal / source
International Conference on Communication, Computing and Electronics Systems (ICCCES)
Pages
1504–1508
Publication date
21 January 2026
Language
EN