Retention prediction in reversed-phase liquid chromatography using XGBoost-based quantitative structure-retention relationships models
Journal of Chromatography A · 2026-09 · 1784 · 467239
DOI: 10.1016/j.chroma.2026.467239 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Chromatography A, 2026-09. Verfasst von Ronald Carrasquillo-Flores und 8 weiteren Autor:innen. Beteiligte Einrichtung: Chemical Process Development, Bristol Myers Squibb, 1 Squibb Dr., New Brunswick, NJ 08903, USA. Electronic address: Ronald.Carrasquillo@BMS.com. Hauptthemen laut MEDLINE: Chromatography, Reverse-Phase, Models, Chemical. Die Arbeit nennt 55 Literaturangaben.
Abstract (Originalsprache)
Reversed-phase liquid chromatography (RPLC) method development remains resource-intensive due to the iterative optimization required to achieve fit-for-purpose separations. While machine learning-enabled quantitative structure-retention relationships (QSRRs) offer predictive capabilities, current models suffer from poor stereoisomer discrimination, uncertain applicability domains, and limited practical integration. This work develops an XGBoost-based QSRR model using 43,329 retention measurements from 86 solutes on 13 stationary phases measured with isocratic mobile phases in the range of 2-90% organic content. The model combines standard 2D Mordred molecular descriptors with 37 custom stereochemical and geometric descriptors that distinguish R/S and E/Z configurations from SMILES strings. Feature selection reduced the descriptor set to 29, achieving a test-set root mean square error (RMSE) of 0.12 (ln selectivity scale), with the addition of custom isomer descriptors providing 2.5-fold improvement in stereoisomer prediction accuracy. The model demonstrated reliable interpolation (RMSE ≤ 0.05, selectivity scale) when molecular descriptor values are within the training-set bounds but showed degraded performance during extrapolation (RMSE 0.11-0.20, selectivity scale). Paired-solute analysis across the 13 stationary phases linked column-dependent prediction failures to descriptors omitted or pruned from the model, (e.g., acid-base character, halogen content, H-bond capacity). This computationally accessible framework provides systematic approaches for assessing prediction reliability across mobile phase compositions and stationary phase chemistries, establishing clear boundaries for trustworthy model application.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Chemical Process Development, Bristol Myers Squibb, 1 Squibb Dr., New Brunswick, NJ 08903, USA. Electronic address: Ronald.Carrasquillo@BMS.com.
- Department of Chemistry, Virginia Commonwealth University, Box 842006, Richmond, VA 23284-2006, USA.
- Department of Chemistry, Gustavus Adolphus College, 800 W. College Ave., St. Peter, MN 56082, USA.
- Chemical Process Development, Bristol Myers Squibb, 1 Squibb Dr., New Brunswick, NJ 08903, USA.
Zitieren
APA
Carrasquillo-Flores, R., Rutan, S. C., Kempen, T., Dahlseid, T., Kruger, Z., Shackman, J. G., Zhou, Y., Wang, Q., & Stoll, D. R. (2026). Retention prediction in reversed-phase liquid chromatography using XGBoost-based quantitative structure-retention relationships models. Journal of Chromatography A, 1784, 467239. https://doi.org/10.1016/j.chroma.2026.467239
Vancouver
Carrasquillo-Flores R, Rutan SC, Kempen T, Dahlseid T, Kruger Z, Shackman JG, et al. Retention prediction in reversed-phase liquid chromatography using XGBoost-based quantitative structure-retention relationships models. Journal of Chromatography A. 2026;1784:467239. doi:10.1016/j.chroma.2026.467239
BibTeX
@article{CarrasquilloFlores2026,
author = {Carrasquillo-Flores, Ronald and Rutan, Sarah C and Kempen, Trevor and Dahlseid, Tina and Kruger, Zachary and Shackman, Jonathan G and Zhou, Yiyang and Wang, Qinggang and Stoll, Dwight R},
title = {Retention prediction in reversed-phase liquid chromatography using XGBoost-based quantitative structure-retention relationships models},
journal = {Journal of Chromatography A},
year = {2026},
volume = {1784},
pages = {467239},
doi = {10.1016/j.chroma.2026.467239}
}
EUVIMED