IodoFinder: Machine Learning-Guided Recognition of Iodinated Chemicals in Nontargeted LC-MS/MS Analysis

Environmental Science & Technology, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 27, 2025

Iodinated disinfection byproducts (I-DBPs) pose significant health concerns due to their high toxicity. Current approaches recognize unknown I-DBPs in mass spectrometry (MS) analysis rely on negative ionization mode, which the characteristic I– fragment can be observed tandem spectra (MS/MS). Still, many ionize exclusively positive where is absent. To address this gap, work developed a machine learning-based strategy iodinated compounds (I-compounds) from MS/MS both electrospray (ESI+) and (ESI−) modes. Investigating over 6000 of 381 I-compounds, we first identified five I-containing neutral losses one diagnostic ESI+ ESI– modes, respectively. We then trained Random Forest models integrated them into IodoFinder, Python program, streamline recognition I-compounds raw LC-MS data. IodoFinder accurately recognized 96% 161 I-compound standards In its application DBP mixtures, discovered 19 with annotated structures an additional 17 assigned formulas, including 12 novel 3 confirmed I-DBPs. envision that will advance identification known exposome studies.

Language: Английский

Dried blood spots analysis for targeted and non-targeted exposomics

medRxiv (Cold Spring Harbor Laboratory), Journal Year: 2025, Volume and Issue: unknown

Published: April 16, 2025

Abstract Dried blood spots (DBS) are an established sample type, widely used in newborn screening programs for monitoring metabolic diseases. Their minimally invasive nature offers great promise assessing chemical exposures, particularly during early life stages and large-scale epidemiological studies. However, comprehensive evaluations of key analytical parameters such as extraction efficiency matrix effects across multiple classes remain limited. Moreover, the promising approach broadly combining targeted non-targeted mass spectrometric data evaluation remains unexplored DBS small-molecule omics. Here, we present optimized LC-HRMS workflow combined exposomic metabolomic analysis samples. Four protocols were systematically compared, with performance evaluated >200 structurally diverse toxicants, pollutants, other biomarkers. The protocol demonstrated acceptable recoveries (60–140%) reproducibility (median RSD: 18%) a majority compounds. Matrix showed median value 76% 14%). In proof-of-principle study, twelve exposure compounds target panel varied physicochemical properties identified real-life samples, several reported first time biomonitoring. Complementary further expanded detectable space, enabling reliable annotation additional exposures. high-confidence identification endogenous metabolites, including amino acids, biogenic amines, fatty acids acylcarnitines capacity to capture broad snapshot human metabolome. These findings support use integrated applications, providing toxicological biological insights from low-volume samples both, prospective retrospective Figure

Language: Английский

MASS SPECTROMETRY-BASED HIGH-THROUGHPUT SAMPLE TREATMENT METHODS FOR ANALYSIS OF XENOBIOTICS IN HUMAN BIOFLUIDS

Advances in Sample Preparation, Journal Year: 2025, Volume and Issue: unknown, P. 100183 - 100183

Published: April 1, 2025

Language: Английский