Neural Network Response Factor Prediction for E&L

The Peer-Reviewed Science Behind Lumo™ 

Accurate quantitation in extractables and leachables (E&L) analysis is complicated by response factor variability, especially in mass spectrometry-based screening methods. This peer-reviewed paper, published in the PDA Journal of Pharmaceutical Science and Technology, introduces in-silico prediction models that help address this challenge.¹ 

What the Paper Covers: 

• The problem: Variability in LC-MS and GC-MS response factors introduces uncertainty into semi-quantitative E&L workflows 
• The approach: Neural network models trained on diverse chemical descriptors, organized into chemistry class-specific sub-models 
• The outcome: High-quality response factor predictions that reduce the need for empirical standards while maintaining defensible accuracy for screening-phase quantitation 

Why It Matters: 
This research forms the scientific foundation for Lumo™, Jordi Labs’ AI-powered response factor prediction capability. The methodology enables faster E&L turnaround, reduced calibration workload, and clearer paths to toxicological risk assessment without sacrificing rigor. 

Reference 
¹ Deng, Y., Grice, A., Louis, M., et al. (2026). Neural Network Prediction of Response Factors for Extractables and Leachables in Pharmaceuticals and Medical Devices. PDA Journal of Pharmaceutical Science and Technology. https://journal.pda.org/content/early/2026/01/30/pdajpst.2025-000061.1

Back to Resources