Ciesielski, D. | bioRxiv Science

bioRxiv2025

ProteoMeter: A pipeline for integrating multi-PTM and limited proteolysis data to reveal modification-structure coupling at the residue level

Systemic perturbations trigger extensive changes across the proteome-altering protein abundance, post-translational modifications (PTMs), conformational states, and complex assembly. Interpreting these effects demands computational pipelines capable of integrating diverse proteomics modalities, such as multi-PTM profiling, limited proteolysis mass spectrometry (LiP-MS), and cross-linking mass spectrometry (XL-MS), within a unified and interoperable framework. Because instrument data are quantified at the peptide level, mapping these measurements to individual residue or modification site is essential for biologically meaningful interpretation. We introduce ProteoMeter, an open-source Python library designed to integrate multi-modal proteomics datasets and map them to single-residue resolution using a standardized coordinate framework. We showcase its capabilities in a combined multi-PTM and LiP-MS analysis profiling the proteomic response to human coronavirus 229e (HCoV-229E) infection.ProteoMeteris actively maintained and is freely available-including all source code and figure-generation scripts-at the following repository: https://github.com/PNNL-Predictive-Phenomics/ProteoMeter.

systems biology↗

bioRxiv2023

QC-GN2oMS2: a Graph Neural Net for High Resolution Mass Spectra Prediction.

Predicting the mass spectrum of a molecular ion is often accomplished via three generalized approaches: rules-based methods for bond breaking, deep learning, or quantum chemical (QC) modeling. Rules-based approaches are often limited by the conditions for different chemical subspaces and perform poorly under chemical regimes with few defined rules. Quantum chemical modeling is theoretically robust but requires significant amounts of computational time to produce a spectrum for a given target. Among deep learning techniques, graph neural networks (GNNs) have performed better than previous work with fingerprint-based neural networks in mass spectral prediction.1 To explore this technique further, we investigate the effects of including quantum chemically derived features as edge features in the GNN to increase predictive accuracy. The models we investigated include categorical bond order, bond force constants derived from Extended Tight-Binding (xTB) quantum chemistry, and acyclic bond dissociation energies. We evaluated these models against a control GNN with no edge features in the input graphs. Bond dissociation enthalpies yielded the best improvement with a cosine similarity score of 0.462 relative to the baseline model (0.437). In this work we also apply dynamic graph attention which improves performance on benchmark problems and supports the inclusion of edge features. Between implementations, we investigate the nature of the molecular embedding for spectral prediction and discuss the recognition of fragment topographies in distinct chemistries for further development in tandem mass spectrometry prediction.

biochemistry↗