Accelerating reaction network identification via frequency-domain analysis
Abstract
The identification of intracellular reaction networks from single-cell data lies at the heart of many biological studies, as it can unravel key mechanisms in living organisms and provide insights for their rational engineering. However, the stochastic and nonlinear nature of these reacting systems poses significant challenges for this identification problem, making the state-of-the-art methods (e.g., particle-filtering-based approaches) computationally demanding. This paper reports a noteworthy numerical observation that for certain systems, the Fourier transform can render the likelihood of the measurement data approximately Gaussian, therefore resulting in a substantially accelerated Bayesian inference algorithm. A gene transcription example is presented to illustrate this finding and demonstrate the efficiency of {the proposed Fourier-based method}.
Related
- The Mean First-Reaction Time on a Thin Filament in a 3-D Domain
- Multi-scale microenvironment covariance for unsupervised spatial domain identification
- FH-Net: Frequency-domain and hierarchical dual enhancement for few-shot surface defect segmentation
- HESTIA: Scalable Multimodal Integration of Histology and High-Resolution Spatial Transcriptomics for Robust Spatial Domain Identification
- Kontext: Kernel-based method bridging spatial domain identification and cell-cell communication
BibTeX
@misc{fang2026accelerating,
title = {{Accelerating reaction network identification via frequency-domain analysis}},
author = {Zhou Fang and Wenying Sun and Xiaoyu Zhang and Mustafa Khammash},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.07684},
eprint = {2611.07684},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.07684},
}
RIS
TY - UNPB
TI - Accelerating reaction network identification via frequency-domain analysis
AU - Zhou Fang
AU - Wenying Sun
AU - Xiaoyu Zhang
AU - Mustafa Khammash
PY - 2026
DA - 2026/11/05
PB - arXiv
DO - 10.48550/arxiv.2611.07684
UR - https://arxiv.org/abs/2611.07684
AB - The identification of intracellular reaction networks from single-cell data lies at the heart of many biological studies, as it can unravel key mechanisms in living organisms and provide insights for their rational engineering. However, the stochastic and nonlinear nature of these reacting systems poses significant challenges for this identification problem, making the state-of-the-art methods (e.g., particle-filtering-based approaches) computationally demanding. This paper reports a noteworthy numerical observation that for certain systems, the Fourier transform can render the likelihood of the measurement data approximately Gaussian, therefore resulting in a substantially accelerated Bayesian inference algorithm. A gene transcription example is presented to illustrate this finding and demonstrate the efficiency of {the proposed Fourier-based method}.
ER -