Intelligent automation of high-performance liquid chromatography method development by means of a real-time knowledge-based approach

• Published in: Journal of Chromatography A 2002-09, Vol.972 (1), p.27-43
• Main Authors: I, Ting-Po, Smith, Randy, Guhan, Sam, Taksen, Ken, Vavra, Mark, Myers, Douglas, Hearn, Milton T.W
• Format: Article
• Language: English
• Subjects: Analytical chemistry; Artificial Intelligence; Automation; Biological and medical sciences; Chemistry; Chromatographic methods and physical methods associated with chromatography; Chromatography, High Pressure Liquid - methods; Exact sciences and technology; General pharmacology; Medical sciences; Other chromatographic methods; Pharmaceutical technology. Pharmaceutical industry; Pharmacology. Drug treatments
• ISSN: 0021-9673
• DOI: 10.1016/S0021-9673(02)01075-0

We describe the development, attributes and capabilities of a novel type of artificial intelligence system, called LabExpert, for automation of HPLC method development. Unlike other computerised method development systems, LabExpert operates in real-time, using an artificial intelligence system and design engine to provide experimental decision outcomes relevant to the optimisation of complex separations as well as the control of the instrumentation, column selection, mobile phase choice and other experimental parameters. LabExpert manages every input parameter to a HPLC data station and evaluates each output parameter of the HPLC data station in real-time as part of its decision process. Based on a combination of inherent and user-defined evaluation criteria, the artificial intelligence system programs use a reasoning process, applying chromatographic principles and acquired experimental observations to iteratively provide a regime for a priori development of an acceptable HPLC separation method. Because remote monitoring and control are also functions of LabExpert, the system allows full-time utilisation of analytical instrumentation and associated laboratory resources. Based on our experience with LabExpert with a wide range of analyte mixtures, this artificial intelligence system consistently identified in a similar or faster time-frame preferred sets of analytical conditions that are equal in resolution, efficiency and throughput to those empirically determined by highly experienced chromatographic scientists. An illustrative example, demonstrating the potential of LabExpert in the process of method development of drug substances, is provided.