Our speakers are Dr. Allen Hubbard, a research scientist in Ivan Baxter’s lab at the Donald Danford Plant Science Center and the CTO of Metablify, and Joshua Ocheje, a graduate student and BP4NTA student member studying under Dr. Natalia Quinete.
Dr. Allen Hubbard will be presenting on The Last Alignment Frontier: Optimal and Non-Monotonic Alignment Across Methods For Signal Amplification.
Abstract: Metablify is an alignment and amplification engine for LC-MS (and other separation + mass spectrometry combinations) data. Metablify provides a solution for LC-MS driven omics analogous to the role that PCR plays in nucleic acid omics. Whereas in genomics, DNA is amplified physically by PCR cycles, with each new sample added to a dataset, the signal for any present chemical is amplified statistically by Metablify’s algorithms. This is particularly valuable for fields such as exposomics where trace levels of a chemical must be detected reliably across samples. The traditional framing of peak picking in individual files, followed by correspondence across them, faces several bottlenecks as dataset size increases, which are exacerbated for low concentration chemicals – i.e. limited signal to noise ratios per sample, pathological shifts, etc. Metablify implements an optimal coordinate transformation between any single sample and the central tendency of a potentially multi-separation method dataset, using dynamic and data derived features. Applying the law of large numbers and novel metrics to harness physical features of the data, Metablify translates the peak correspondence challenge into the language of optimal transport. Doing so, Metablify is unrestrained by the limitations of traditional alignment approaches such as dynamic time warping (which assumes monotonic drift). Metablify enables alignment of signals up until the point of peak reversal, a situation that would previously be hopelessly unresolvable. Metablify can drive untargeted metabolomics or proteomics, and could be extremely valuable to exposomics and related fields.
Joshua Ocheje will be presenting on Leachability and Chemical Profiles of Per- and Polyfluoroalkyl Substances in Electronic Waste Components: Targeted and Non-Targeted Analysis.
Abstract: Electronic waste (e-waste) is a growing solid waste stream with largely undisclosed and poorly characterized fluorinated constituents. We evaluated per- and polyfluoroalkyl substances (PFAS) leachability from four e-waste components (phone screens, phone plastics, capacitors, and Lithium-ion batteries) using a 30-day deionized water leaching test. PFAS were extracted by solid-phase extraction using weak anion exchange (WAX) cartridges and analyzed with a liquid chromatography triple-quadrupole mass spectrometer. In addition, the PFAS chemical profiles of e-waste components were characterized by non-targeted analysis. Leachable sums of detected PFAS (∑PFAS) were highest in phone screens (1739–1932 ng·kg−1) and phone plastics (1575–2197 ng·kg−1) and an order of magnitude lower in Lithium-ion batteries (148–158 ng·kg−1) and capacitors (147–243 ng·kg−1). Short-chain perfluoroalkyl acids (PFAAs) (e.g., PFBA, PFHxA) and legacy acids
(e.g., PFOA, PFNA) were more prevalent in phone screens/plastics, whereas capacitors and batteries showed mixed sulfonate/carboxylate patterns (PFOS, PFHxS, and 6:2 FTS). Although capacitors and Lithium-ion batteries contained essential PFAS with high hazard potential at trace levels, phone screens and phone plastics pose a greater risk per mass due to higher ∑PFAS levels and larger volumes. Non-targeted analysis using Orbitrap Astral revealed CF2/CF2O homologous trends with corroborating targeted findings. These findings highlight the need for PFAS-free alternatives, the disclosure of fluorinated additives, and stronger end-of-life management strategies to prevent PFAS releases from e-waste.