Prediction of Collision Cross-Section Values for Extractables and Leachables from Plastic Products

Abstract

The use of ion mobility separation (IMS) in conjunction with high-resolution mass spectrometry has proved to be a reliable and useful technique for the characterization of small molecules from plastic products. Collision cross-section (CCS) values derived from IMS can be used as a structural descriptor to aid compound identification. One limitation of the application of IMS to the identification of chemicals from plastics is the lack of published empirical CCS values. As such, machine learning techniques can provide an alternative approach by generating predicted CCS values. Herein, experimental CCS values for over a thousand chemicals associated with plastics were collected from the literature and used to develop an accurate CCS prediction model for extractables and leachables from plastic products. The effect of different molecular descriptors and machine learning algorithms on the model performance were assessed. A support vector machine (SVM) model, based on Chemistry Development Kit (CDK) descriptors, provided the most accurate prediction with 93.3% of CCS values for [M + H] + adducts and 95.0% of CCS values for [M + Na] + adducts in testing sets predicted with <5% error. Median relative errors for the CCS values of the [M + H] + and [M + Na] + adducts were 1.42 and 1.76%, respectively. Subsequently, CCS values for the compounds in the Chemicals associated with Plastic Packaging Database and the Food Contact Chemicals Database were predicted using the SVM model developed herein. These values were integrated in our structural elucidation workflow and applied to the identification of plastic-related chemicals in river water. False positives were reduced, and the identification confidence level was improved by the incorporation of predicted CCS values in the suspect screening workflow.

Identification of Nonvolatile Migrates from Food Contact Materials Using Ion Mobility–High-Resolution Mass Spectrometry and in Silico Prediction Tools

“…In the CCS prediction model, the CCS values of 93.3% of [M + H] + adducts and 95.0% of [M + Na] + adducts in the test data set were predicted with less than 5% error. 37 Therefore, the tolerance used for the CCS values on screening the measured data against the CPPdb and FCCdb libraries was set to 5%.…”

“…The prediction of CCS values of chemicals associated with plastic products has been described in a previous study, in which the CCS prediction models were built using support vector machine regression (SVM) based on 1076 [M + H] + CCS values and 645 [M + Na] + CCS values. As CCS values are reproducible across different laboratories and platforms, some CCS records for model building were from other publications. ,,− …”

“…19,25,38−41 Prediction of RT and CCS Values of Substances in CPPdb and FCCdb. The rationalization of the substances in the CPPdb and FCCdb databases was performed in a previous study, 37 where metals and salts, together with any substances with the same InChIKey (replicates), were removed. Additionally, compounds with a molecular weight outside the range 50−1200 Da were removed, leaving 2883 and 6508 substances in the CPPdb and FCCdb, respectively.…”

Count-Based Morgan Fingerprint: A More Efficient and Interpretable Molecular Representation in Developing Machine Learning-Based Predictive Regression Models for Water Contaminants’ Activities and Properties

“…p K d and pIC 50 measure the binding affinity of PPARγ, a ligand-activated nuclear receptor involved in multifaceted physiology and chemically induced endocrine disruption . CCS, derived from ion mobility separation (IMS), is a physicochemical property of ions and is related to the chemical structure and three-dimensional conformation of the molecules . FreeSolv provides the experimental hydration free energy of small molecules in water, while Lipo offers experimental results of the octanol/water distribution coefficient (log D at pH 7.4) of 4200 compounds .…”

Collision cross section measurement and prediction methods in omics

“…170 Despite the higher prediction error found for some compound classes, according to da Silva et al, it is suggested to use AllCCS as the primary tool for confirmation because it has the highest coverage in CCS prediction compared with other tools and has a perfect Pearson correlation coefficient between experimental and theoretical CCSs. 171 Moreover, it can also provide CCS predictions for small molecules such as drugs, 105 natural products, 172 food contact materials, 155,156 and pesticides. Two clusters contained small molecules, one cluster had lipids associated with it, and the last cluster consisted of carbohydrates and peptides.…”