Uncertainty-Aware Learning of Multiple Conditions as a Framework for Streamlined Retention Time Prediction to Accelerate Method Development

Liquid chromatography is a cornerstone analytical technique for separating, quantifying, and purifying components from complex mixtures. To accelerate method screening, we herein introduce an uncertainty-aware graph-based neural network that predicts retention times across multiple column chemistries and buffer pH conditions. Specifically, the multiple condition retention time model, MC-Retention, incorporates explicit column and buffer descriptors and is trained on a newly assembled data set produced by screening 480 analytes under eight chromatography conditions. By being able to accurately model multiple chromatographic methods simultaneously, MC-Retention can substantially reduce the time and resources required for LC method screening and development. With an aggregate <i>R</i>² of 0.86 and a mean absolute error (MAE) of 15.5 s during cross-validation, MC-Retention's respectable performance is greatly enhanced for analytes made present during training for one of the four column chemistries, when those analytes are otherwise not associated with the columns being validated. This selective training of analytes for a single column chemistry reduces error by 50%, with an <i>R</i>² of 0.95 and MAE of 8.2 s in aggregate. Additionally, we demonstrate that MC-Retention effectively identifies optimal conditions for separating amide bond-forming reaction components, while also supplying calibrated uncertainty estimates for its retention time predictions.