msengine: Adaptive Open-Source Signal Extraction for HDX MS
- 1Genentech
Abstract
Hydrogen-deuterium exchange mass spectrometry (HDX-MS) requires extraction of broad, time-dependent peptide isotope envelopes from noise and overlapping signals. The ExMS/ExMS2 lineage established a retention-time-aware framework for raw-data extraction and integrated HDX-MS analysis. Building on this foundation, msengine is an open-source Python engine that adapts mass-selection windows and retention-time boundaries to the observed data. Spectra are reconstructed as nonnegative combinations of the theoretical undeuterated (all-H) isotope distribution shifted across allowed deuteration states, without prescribing a population shape. The fitted weights, termed deuteration coefficients, describe each state's contribution and guide refinement of mass-selection windows. Spectral matches establish an initial retention interval, which is then expanded using chromatographic intensity to include weaker portions of the elution profile. Interference screening and final quality checks assess the recovered interval. In a 1:1 mixture of undeuterated and deuterated material, interval expansion changed the relative population intensities within the extracted envelope resulting in closer agreement to the 1:1 target, illustrating the importance of boundary selection. Across a four-timepoint experiment with the SET1/SWD2 complex collected in duplicate, msengine accepted an average of 901 peptide-charge targets per measurement versus 636 for ExMS2, a 41.7% increase, while improving replicate-centroid disagreement and retention-window widths between replicates. In a same-workstation benchmark, total wall time decreased from 71.9 to 7.3 min, with higher peak memory use. Together, these results demonstrate adaptive signal extraction across mass and retention time as a means to retain multimodal envelopes while increasing accepted-target yield, persistence, accuracy, and throughput.
Related
- From Signal to Memory: A Proposed Adaptive Hierarchy Linking Autism-Associated Genes to Sensory Homeostasis
- DBCLS BioHackathon 2026 report: Open, Reproducible Extraction of Data-Entity References from the Biomedical Literature – with Entry-Level Resolution via Identifiers.org / TogoID
- Pruning the Search, Not the Signal: Adaptive-Banding Needleman-Wunsch Sequence Alignment via Protein Language Model Confidence
- SkeleSketch: An Open-Source Tool for Quantifying Microglial Morphology in Fiji/ImageJ
- An open-source platform for measuring optomotor responses in mice
BibTeX
@misc{patapoff2026msengine,
title = {{msengine: Adaptive Open-Source Signal Extraction for HDX MS}},
author = {A. Patapoff and N. Patapoff and B. T. Walters},
howpublished = {bioRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.30.755802},
url = {https://www.biorxiv.org/content/10.64898/2026.09.30.755802v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - msengine: Adaptive Open-Source Signal Extraction for HDX MS
AU - A. Patapoff
AU - N. Patapoff
AU - B. T. Walters
PY - 2026
DA - 2026/09/30
PB - bioRxiv
DO - 10.64898/2026.09.30.755802
UR - https://www.biorxiv.org/content/10.64898/2026.09.30.755802v1
AB - Hydrogen-deuterium exchange mass spectrometry (HDX-MS) requires extraction of broad, time-dependent peptide isotope envelopes from noise and overlapping signals. The ExMS/ExMS2 lineage established a retention-time-aware framework for raw-data extraction and integrated HDX-MS analysis. Building on this foundation, msengine is an open-source Python engine that adapts mass-selection windows and retention-time boundaries to the observed data. Spectra are reconstructed as nonnegative combinations of the theoretical undeuterated (all-H) isotope distribution shifted across allowed deuteration states, without prescribing a population shape. The fitted weights, termed deuteration coefficients, describe each state's contribution and guide refinement of mass-selection windows. Spectral matches establish an initial retention interval, which is then expanded using chromatographic intensity to include weaker portions of the elution profile. Interference screening and final quality checks assess the recovered interval. In a 1:1 mixture of undeuterated and deuterated material, interval expansion changed the relative population intensities within the extracted envelope resulting in closer agreement to the 1:1 target, illustrating the importance of boundary selection. Across a four-timepoint experiment with the SET1/SWD2 complex collected in duplicate, msengine accepted an average of 901 peptide-charge targets per measurement versus 636 for ExMS2, a 41.7% increase, while improving replicate-centroid disagreement and retention-window widths between replicates. In a same-workstation benchmark, total wall time decreased from 71.9 to 7.3 min, with higher peak memory use. Together, these results demonstrate adaptive signal extraction across mass and retention time as a means to retain multimodal envelopes while increasing accepted-target yield, persistence, accuracy, and throughput.
ER -