隨著科技日益發展，質譜儀（MS）與核磁共振光譜儀（NMR）等高通量技術已能夠在短時間內產生大量且高解析度的代謝體資料。代謝體學作為系統生物學的重要一環，能夠提供細胞代謝狀態的即時資訊，對於疾病機制研究、生物標記發掘及精準醫療發展皆具有重要意義。在以質譜儀為基礎的代謝體定量技術中，液相層析串聯質譜儀（LC-MS）具有高靈敏度以及可偵測大量代謝物的特性，因而廣泛應用於代謝體研究中。然而，LC-MS代謝體資料通常會有批次效應，使得後續的資料分析充滿挑戰。因此，本研究將透過模擬研究觀察ComBat和Limma二種批次效應校正方法應用於LC-MS代謝體資料在偏最小平方法判別模型（PLS-DA）的表現，其模型評估指標包含準確度以及1-特異度，接著進一步探討處理效應與批次效應對代謝體分析的影響，並提供研究者代謝體分析上的建議。

Metabolomics, as an integral part of systems biology, provides real-time insights into cellular metabolic states and plays a pivotal role in elucidating disease mechanisms, discovering biomarkers, and driving the development of precision medicine. With the recent advances in high-throughput technologies, such as Mass Spectrometry (MS) and Nuclear Magnetic Resonance (NMR) the generation of large-scale, high-resolution metabolomics data has been greatly facilitated. Among MS-based strategies, Liquid Chromatography-Mass Spectrometry (LC-MS) is widely used for its high sensitivity and broad metabolite coverage, enabling accurate detection and quantification of diverse metabolites. However, despite these analytical advantages, LC-MS-based metabolomics data are often affected by batch effects, which can obscure true biological signals and hinder downstream analysis.
Therefore, our research employs simulation studies to assess the effectiveness of two batch effect correction methods—ComBat and Limma—applied to LC-MS-based metabolomics data analyzed using the Partial Least Squares Discriminant Analysis (PLS-DA) model. The performance of each method will be evaluated in terms of classification accuracy and 1-specificity. In addition, we will examine the impact of both treatment effects and batch effects on metabolomics data, aiming to provide practical guidance for researchers working in this field.

Burankova, Y., Klemm, J., Lohmann, J. J. G., Taheri, A., Probul, N., Baumbach, J. and Zolotareva, O. (2024). FedRBE – a decentralized privacy-preserving federated batch effect correction tool for omics data based on limma, arXiv e-prints p. arXiv:2412.05894.
Chong, I.-G. and Jun, C.-H. (2004). Performance of some variable selection methods when multicollinearity is present, Chemometrics and Intelligent Laboratory Systems 78(1–2): 103–12.
Han, W. and Li, L. (2020). Evaluating and minimizing batch effects in metabolomics, Mass Spectrometry Reviews 41(3): 421–42.
Johnson, W. E., Li, C. and Rabinovic, A. (2007). Adjusting batch effects in microarray expression data using empirical bayes methods, Biostatistics 8(1): 118–27.
Kim, T., Tang, O., Vernon, S. T., Kott, K. A., Koay, Y. C., Park, J., James, D. E., Grieve, S. M., Speed, T. P., Yang, P., Figtree, G. A., O’Sullivan, J. F. and Yang, J. Y. H. (2021). A hierarchical approach to removal of unwanted variation for large-scale metabolomics data, Nature Communications 12(4992).
Leek, J. T., Scharpf, R. B., Bravo, H. C., Simcha, D., Langmead, B., Johnson, W. E., Geman, D., Baggerly, K. and Irizarry, R. A. (2010). Tackling the widespread and critical impact of batch effects in high-throughput data, Nature Reviews Genetics 11: 733–39.
Nakamura, N., Pence, L. M., Cao, Z. and Beger, R. D. (2020). Distinct lipid signatures are identified in the plasma of rats with chronic inflammation induced by estradiol benzoate and sex hormones, Metabolomics 16(95).
Renard, E. and Absil, P.-A. (2017). Comparison of location-scale and matrix factorization batch effect removal methods on gene expression datasets, 2017 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) pp. 1530–37.
Ye, H., Zhang, X., Wang, C., Goode, E. L. and Chen, J. (2023). Batch-effect correction with sample remeasurement in highly confounded case-control studies, Nature Computational Science 3: 709–19.