The Mo Fanyang Research Group of the School of Materials Science and Engineering has made progress in conducting cross-cutting research on AI for Science

Thin layer chromatography (TLC) is widely used in synthetic chemistry laboratories and is one of the most commonly used analytical methods for the monitoring of organic reactions as well as for determining the location

Mo Fanyang's research group of the School of Materials Science and Engineering built a robot platform, developed the automation technology of TLC analysis, obtained a large amount of standardized TLC data, and then applied machine learning to regression analysis of the data to obtain a model

Automation and machine learning are combined to establish TLC predictive models

The researchers used MACCSKey molecular fingerprints and physicochemical descriptors, such as molecular weight (MW), topological polarity surface area (TPSA), and hydrogen bond donor number (HBD), to convert molecules into numerical structures that computers can process; Weighted vector coding is used to represent mobile phase information

Feature engineering and model prediction accuracy

The researchers assessed the relative importance

The relevant research results were published in Chem (Doi:10.

Related video links:

Links:

https://doi.