Gene Discovery in MALT Lymphoma Using Explainable Machine Learning on scRNA-seq Data

Abstract

Mucosa-associated lymphoid tissue (MALT) lymphoma is an indolent and clinically heterogeneous B-cell malignancy. The lack of robust molecular biomarkers and the limited availability of high-resolution single-cell studies hinder the development of precise diagnostic and therapeutic strategies. To address this gap, an interpretable machine learning pipeline was developed and applied to single-cell RNA sequencing (scRNA-seq) data derived from MALT lymphoma patients and healthy donors. After quality control, normalization, and dimensionality reduction via truncated singular value decomposition (SVD), multiple tree-based classifiers were trained using stratified cross-validation and evaluated on a held-out balanced test set. Feature importance scores and SHapley Additive exPlanations (SHAP) values were integrated with Wilcoxon rank-sum testing to identify statistically supported gene markers. Among the classifiers, CatBoost, LightGBM, and Random Forest achieved the highest performance. Genes such as RPS4Y1, RGS1, XIST, CREM, and HSPH1 were consistently prioritized by both SHAP and statistical testing, indicating their biological relevance and differential expression. Notably, a divergence was observed between impurity-based feature importance rankings and the SHAP/Wilcoxon consensus, reflecting the complementary nature of these analytical approaches. This integrative framework provides a transparent and reproducible approach for gene discovery in scRNA-seq data and contributes computationally prioritized candidate genes warranting experimental validation for understanding MALT lymphoma pathogenesis and improving future molecular diagnostics.

Keywords

• scRNA-seq
• MALT lymphoma
• Explainable machine learning
• SHAP
• Biomarker discovery
• Feature importance

References

1. Andor, N., Simonds, E. F., Czerwinski, D. K., Chen, J., Grimes, S. M., Wood-Bouwens, C., Ji, H. P., & Levy, R. (2019). Single-cell RNA-Seq of follicular lymphoma reveals malignant B-cell types and coexpression of T-cell immune checkpoints. Blood, 134(13), 1119–1129. https://doi.org/10.1182/blood-2018-08-862292
2. Bobée, V., Stamatoullas, A., Lenain, P., Parmentier, F., Camus, V., Picquenot, J. M., & Tilly, H. (2020). Combining gene expression profiling and machine learning to diagnose B-cell non-Hodgkin lymphoma. Leukemia & Lymphoma, 61(4), 902–911. https://doi.org/10.1038/s41408-020-0322-5
3. Boye, K., & Maelandsmo, G. M. (2010). S100A4 and metastasis: A small actor playing many roles. The American Journal of Pathology, 176(2), 528–535. https://doi.org/10.2353/ajpath.2010.090526
4. Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 785–794). https://doi.org/10.1145/2939672.2939785
5. Doherty, L., Sheen, M. R., Vlachos, A., Choesmel, V., O’Donohue, M. F., Clinton, C., et al. (2010). Ribosomal protein genes RPS10 and RPS26 are commonly mutated in Diamond-Blackfan anemia. American Journal of Human Genetics, 86(2), 222–228. https://doi.org/10.1016/j.ajhg.2009.12.015
6. Genomics-10x. (2024). Chromium single cell gene expression v3 data. https://www.10xgenomics.com/ (accessed July 1, 2024)
7. Ilicic, T., Kim, J. K., Kolodziejczyk, A. A., et al. (2016). Classification of low quality cells from single-cell RNA-seq data. Genome Biology, 17, 29. https://doi.org/10.1186/s13059-016-0888-1
8. Juang, Y. T., Wang, Y., Solomou, E. E., Li, Y., Mawrin, C., Tenbrock, K., Kyttaris, V. C., & Tsokos, G. C. (2005). Systemic lupus erythematosus serum IgG increases CREM binding to the IL-2 promoter and suppresses IL-2 production through CaMKIV. Journal of Clinical Investigation, 115(4), 996–1005. https://doi.org/10.1172/JCI22854

9. Lenz, G., Wright, G., Dave, S. S., Xiao, W., Powell, J., Zhao, H., et al. (2008). Stromal gene signatures in large-B-cell lymphomas. The New England Journal of Medicine, 359(22), 2313–2323. https://doi.org/10.1056/NEJMoa0802885
10. Liu, J., Zhang, X., Lin, M., Chen, L., Chen, Y., & Wang, Y. (2019). ABI3 downregulation is associated with lymph node metastasis and poor prognosis in papillary thyroid carcinoma. Pathology Research and Practice, 215(10), 152592. https://doi.org/10.1016/j.prp.2019.152592
11. Lopes-Ramos, C. M., Quackenbush, J., & DeMeo, D. L. (2020). Genome-wide sex and gender differences in gene expression. Nature Communications, 11, 1395. https://doi.org/10.1038/s41467-020-19076-2
12. Luecken, M. D., & Theis, F. J. (2019). Current best practices in single‐cell RNA‐seq analysis: A tutorial. Molecular Systems Biology, 15(6), Article e8746. https://doi.org/10.15252/msb.20188746
13. Lundberg, S. M., & Lee, S. I. (2017). A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 30.
14. Moratz, C., Hayman, J. R., Gu, H., & Kehrl, J. H. (2004). Abnormal B-cell responses to chemokines, disturbed plasma cell localization, and distorted immune tissue architecture in Rgs1−/− mice. Molecular and Cellular Biology, 24(13), 5767–5775. https://doi.org/10.1128/MCB.24.13.5767-5775.2004
15. Pore, D., Bodo, J., Danda, A., Yan, D., Phillips, J. G., Lindner, D., et al. (2015). Identification of Ezrin-Radixin-Moesin proteins as novel regulators of pathogenic B-cell receptor signaling and tumor growth in diffuse large B-cell lymphoma. Leukemia, 29, 1857–1867. https://doi.org/10.1038/leu.2015.107
16. Rodríguez-Sevilla, J., & Salar, A. (2022). Recent advances in the genetics of MALT lymphomas. Cancers, 14(1), 176. https://doi.org/10.3390/cancers14010176
17. Van Rossum, G., & Drake, F. L. (2009). Python 3 reference manual. CreateSpace.
18. Wang, J., Zhang, Y., Zhang, W., Li, Y., Hu, X., Li, X., et al. (2021). A comprehensive analysis of long noncoding RNAs in colorectal cancer reveals novel biomarkers and potential therapeutic targets. Molecular Medicine Reports, 23(4), 296. https://doi.org/10.3892/mmr.2021.11935
19. Weitzman, J. B., Fiette, L., Matsuo, K., & Yaniv, M. (2000). JunD protects cells from p53-dependent senescence and apoptosis. Molecular and Cellular Biology, 20(2), 593–602. https://doi.org/10.1128/MCB.20.2.593-602.2000
20. Yildirim, E., Kirby, J. E., Brown, D. E., Mercier, F. E., Sadreyev, R. I., Scadden, D. T., & Lee, J. T. (2013). Xist RNA is a potent suppressor of hematologic cancer in mice. Cell, 152(4), 727–742. https://doi.org/10.1016/j.cell.2013.01.034
21. Zappasodi, R., Ruggiero, G., Guarnotta, C., Tortoreto, M., Tringali, C., Cavane, A., et al. (2015). HSPH1 inhibition downregulates Bcl-6 and c-Myc and hampers the growth of human aggressive B-cell non-Hodgkin lymphoma. Blood, 125(13), Article 1768–1771. https://doi.org/10.1182/blood-2014-09-599308
22. Zhang, Y., Li, X., Liu, H., Wang, Z., Chen, Y., Huang, Z., et al. (2025). Single-cell RNA-sequencing reveals cellular heterogeneity and immune microenvironment in ocular adnexal MALT lymphoma. Frontiers in Immunology, 16, 1508559. https://doi.org/10.3389/fimmu.2025.1508559
23. Zhao, X., Liu, L., Liu, D., Zhang, Y., Zhang, J., Zhang, Q., et al. (2021). Prediction of survival and immunotherapy response by immune cell infiltration score in diffuse large B-cell lymphoma. Aging, 13(3), 3400–3416. https://doi.org/10.18632/aging.202431