Capillary electrophoresis - unnecessary due to AI?

Hermann Wätzig

In recent years, there have been numerous advances and new developments in the field of capillary electrophoresis (CE). The pros and cons of the various instruments are becoming increasingly well understood, and the rules for method development have been clarified, particularly with regard to larger molecules, including biopharmaceuticals. CE-MS has evolved significantly and is now increasingly becoming a routine option for quality control. Last but not least, affinity CE assays complement the range of instruments available for binding assays.

Fundamentals form the basis of every science. For CE, this foundation was largely laid by Czech scientists. This has resulted in numerous publications, and to highlight one in particular: the CEval software has established itself as the standard tool worldwide. “Artificial intelligence” will not replace clever algorithmic solutions. Algorithmic solutions will remain important, but in some cases will be improved by AI. AI holds many promises – and as always, promises are sometimes kept.

Peak integration in CE remains a challenge. There are numerous ideas for improving peak integration, ranging from simple algorithms for improving the signal-to-noise ratio to more complex AI solutions. However, none of these ideas can prove that they are genuine improvements, as they cannot be compared with each other or with previous algorithms using the same reference data sets. Therefore, data sets were collected, standardized, and then integrated by teams of experts. This consolidated data will soon be available to the scientific community.

So, to answer the question in the title: NO! However, when used sensibly, AI can significantly improve science in many areas, including CE.