Structure-based prediction of reversed-phase liquid chromatography retention time using machine-learning models built from miniature datasets
Analytica Chimica Acta · 2026-12 · 1428 · 346198
DOI: 10.1016/j.aca.2026.346198 ↗Kurz gefasst
Fachartikel in der Zeitschrift Analytica Chimica Acta, 2026-12. Verfasst von Jane Kawakami und 8 weiteren Autor:innen. Beteiligte Einrichtung: Analytical Research & Development, Pfizer Inc., Eastern Point Road, Groton, CT, 06340, USA. Electronic address: Jie.Kawakami@pfizer.com. Hauptthemen laut MEDLINE: Chromatography, Reverse-Phase, Machine Learning. Die Arbeit nennt 57 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
: Structure-based chromatographic retention prediction is critical for efficient method development and impurity characterization or monitoring in pharmaceutical analysis yet remains challenging without proper databases. Here, we present a practical project-specific quantitative structure–retention relationship (QSRR) framework that integrates descriptor fusion with a rigorously designed feature- and model-selection strategy to enable accurate retention time (RT) prediction using unfiltered miniature reversed-phase liquid chromatography (RPLC) datasets. Six molecular descriptor libraries, along with their various combinations, were evaluated using multiple machine learning algorithms within a nested cross-validation (CV) framework optimized to balance model fit and generalizability. The workflow was benchmarked using three late-stage Pfizer small-molecule projects, each comprising 17–21 compounds, including an externally validated set. High prediction accuracy was achieved across all case studies (R 2 > 0.96, mean absolute percentage error < 9%), with absolute RT errors below 2.5 minutes (for analysis time ranging between 30 and 60 min) for all compounds in the EV set, including regioisomeric impurities. Notably, unlike the conventional localized QSRR modeling, the project-specific QSRR modeling does not require explicit structural similarity filtering because the compounds for model development and external validation are inherently chemically related.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Substanzen
Autor:innen und Institutionen
- Analytical Research & Development, Pfizer Inc., Eastern Point Road, Groton, CT, 06340, USA. Electronic address: Jie.Kawakami@pfizer.com.
- Portfolio Statistics, PharmSci Transformational Technology, Pfizer Inc., Eastern Point Road, Groton, CT, 06340, USA.
- Analytical Research & Development, Pfizer Inc., Eastern Point Road, Groton, CT, 06340, USA.
- Medicine Design, Pfizer Inc., Eastern Point Road, Groton, CT, 06340, USA.
- Information Management, PharmSci Transformational Technology, Pfizer Inc., Eastern Point Road, Groton, CT, 06340, USA.
- Chemical Research & Development, Pfizer Inc., Eastern Point Road, Groton, CT, 06340, USA.
Zitieren
APA
Kawakami, J., Khan, M. K. H., Liu, Y., Dybeck, E., Farrand, D., Du, B., Mustakis, I., Zhang, Y., & Wang, K. (2026). Structure-based prediction of reversed-phase liquid chromatography retention time using machine-learning models built from miniature datasets. Analytica Chimica Acta, 1428, 346198. https://doi.org/10.1016/j.aca.2026.346198
Vancouver
Kawakami J, Khan MKH, Liu Y, Dybeck E, Farrand D, Du B, et al. Structure-based prediction of reversed-phase liquid chromatography retention time using machine-learning models built from miniature datasets. Analytica Chimica Acta. 2026;1428:346198. doi:10.1016/j.aca.2026.346198
BibTeX
@article{Kawakami2026,
author = {Kawakami, Jane and Khan, Md Kamrul Hasan and Liu, Yizhou and Dybeck, Eric and Farrand, Doug and Du, Bo and Mustakis, Iason and Zhang, Yu and Wang, Ke},
title = {Structure-based prediction of reversed-phase liquid chromatography retention time using machine-learning models built from miniature datasets},
journal = {Analytica Chimica Acta},
year = {2026},
volume = {1428},
pages = {346198},
doi = {10.1016/j.aca.2026.346198}
}
EUVIMED