Wednesday 19 November 2025
The rapid growth of mass spectrometry data has created a significant challenge for efficient library search, a crucial step in drug discovery and biomarker identification. Traditional methods struggle to keep up with the vast amounts of data, leading to inaccurate results and lengthy processing times. To address this issue, researchers have turned to in-storage processing (ISP) techniques that utilize specialized memory devices to accelerate computations.
A recent study has introduced an innovative ISP architecture that leverages a 3D Ferroelectric NAND (FeNAND) structure to enhance the speed and efficiency of mass spectrometry library search. The FeNAND device offers significantly higher density, faster speeds, and lower voltage requirements compared to traditional NAND flash memory.
The researchers combined the FeNAND device with hyperdimensional computing (HDC), a brain-inspired paradigm that enables highly parallel processing with simple operations and strong error tolerance. This integration allows for efficient matching of experimental spectra to reference libraries, even when novel peptides or post-translational modifications are present.
To further optimize performance, the team developed a dual-bound approximate matching (D-BAM) distance metric tailored to the FeNAND structure. This metric enables parallel vector computations that significantly accelerate the library search process.
The resulting ISP architecture achieved impressive results, with a 43-fold speedup and 21-fold increase in energy efficiency compared to state-of-the-art 3D NAND methods. Moreover, the accuracy of the results remained comparable to traditional methods.
The implications of this research are significant for various fields, including proteomics and metabolomics. By accelerating mass spectrometry library search, researchers can quickly identify biomarkers associated with diseases and develop more effective treatments. The FeNAND-based ISP architecture also has potential applications in other areas where high-performance computing is crucial, such as artificial intelligence and machine learning.
The development of this innovative ISP architecture highlights the importance of interdisciplinary collaboration between experts from various fields, including computer science, materials science, and biology. By combining their expertise, researchers can create novel solutions that tackle complex challenges and push the boundaries of what is possible.
In the future, it will be exciting to see how this technology is further developed and applied to real-world problems. As the field continues to evolve, we can expect even more innovative solutions that harness the power of FeNAND and HDC to accelerate scientific discovery and improve our understanding of the world around us.
Cite this article: “Accelerating Mass Spectrometry Library Search with In-Storage Processing and Hyperdimensional Computing”, The Science Archive, 2025.
Mass Spectrometry, Library Search, In-Storage Processing, 3D Ferroelectric Nand, Hyperdimensional Computing, Parallel Processing, Distance Metric, Energy Efficiency, Proteomics, Metabolomics.
