Advances in artificial intelligence and machine learning in capillary electrophoresis

Abstract

Capillary electrophoresis (CE) is considered a good separation technique due to its inexpensiveness and speed, but it suffers from various drawbacks, like reproducibility and method development challenges. These problems are being tackled by integrating CE with artificial intelligence (AI) and machine learning (ML) tools. Some software and models have been used in CE for achieving better performance, but the integration is still in progress. In this article, an effort was made to provide a complete status of CE integration with AI and ML approaches. Various aspects of AI and ML integration with CE have been discussed, including types of software and models utilized, method development and optimization approaches, peak recognition and deconvolution, retention time forecast, data analysis, signal correction, analyte documentation and quality control with predictive maintenance. Also, the uses of AI- and ML-integrated CE in real-life samples were addressed. Integration of AI and ML into nano-CE has also been discussed. Furthermore, a comparison between conventional and AI-based CE was carried out. Finally, the challenges, future perspectives and recommendations were discussed to make the integration highly useful in the near future for a wide range of applications.