Stream-Based Compression of Mass Spectrometry Signal via Spectral Access Units and Magnitude-Stratified Predictive Coding
- 1 May 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33 (10709908) , 1816-1820
Abstract
High-throughput mass spectrometry generates massive non-stationary 1D signals, posing challenges to storage efficiency and interactive retrieval. Current monolithic compression schemes impose random access barriers, hindering parallel processing. To address this, this paper proposes a stream-based compression framework governed by a Spectral Access Unit (SAU) architecture, which encapsulates signal segments into self-contained packets for $O(1)$ retrieval and robust error containment. At the algorithmic level, we introduce Magnitude-Stratified Predictive Coding (MSPC), which extracts global Most Probable Modes (MPMs) at specific magnitude strata as prediction priors. By applying Gray-mapped Boolean masking between significant digits and their MPMs, MSPC leverages the adjacency of Gray-code space to transform minor numerical deviations into highly sparse, zero-centric residuals. Coupled with an Original Domain-Based Logarithmic Binary Arithmetic Coding (ODB-LBAC) engine, this effectively eliminates structural redundancies. Experimental results demonstrate that the proposed PMSDC achieves superior compression performance and accelerated encoding throughput compared to established general-purpose and domain-specific baselines, all while strictly preserving high signal fidelity for downstream analytics. The proposed PMSDC has been adopted by the Audio Video Coding Standard Workgroup of China (AVS) as the standard for mass spectrometry data compression.Keywords
Funding Information
- Fundamental and Interdisciplinary Disciplines Breakthrough Plan of the Ministry of Education of China (JYB2025XDXM413)
This publication has 27 references indexed in Scilit:
- Dear-OMG: An Omics-General Compression Method for Genomics, Proteomics and Metabolomics DataInterdisciplinary Sciences: Computational Life Sciences, 2025
- Aird-MSI: A High Compression Rate and Decompression Speed Format for Mass Spectrometry Imaging DataJournal of Proteome Research, 2025
- An Advancement in Huffman Coding With a Potential for Parallel DecodingConcurrency and Computation: Practice and Experience, 2025
- How Much Storage Precision Can Be Lost: Guidance for Near-Lossless Compression of Untargeted Metabolomics Mass Spectrometry DataJournal of Proteome Research, 2024
- Improved Mass Accuracy and Precision for Multi-Attribute Methods Using a New Internally Calibrated High Resolution Orbitrap Mass DetectorAnalytical Chemistry, 2024
- Acquisition and Analysis of DIA-Based Proteomic Data: A Comprehensive Survey in 2023Molecular & Cellular Proteomics, 2024
- LDA enhanced one-bit compressive sensing method for high-throughput mass spectrometry data feature selectionDigital Signal Processing, 2023
- A faster implementation of online RLBWT and its application to LZ77 parsingJournal of Discrete Algorithms, 2018
- Mass Accuracy and Isotopic Abundance Measurements for HR-MS Instrumentation: Capabilities for Non-Targeted AnalysesJournal of the American Society for Mass Spectrometry, 2014
- mzML: A single, unifying data format for mass spectrometer outputProteomics, 2008