Early diagnosis of gastric cancer is critical for improved patient prognosis. MicroRNAs (miRNAs), a class of non-coding small molecules (20-25 nucleotides) that regulate gene expression by binding to target RNAs, represent promising disease biomarkers due to their inherent stability in bodily fluids. In this study, based on a dataset of 2,834 serum samples sourced from the Gene Expression Omnibus (GEO) database (1,417 early gastric cancer patients and 1,417 healthy controls), four penalized logistic regression models—LASSO, Elastic Net (ENet), Smoothly Clipped Absolute Deviation (SCAD), and Minimax Concave Penalty (MCP)—are employed for feature selection. These models are subsequently integrated with a coordinate descent algorithm to develop a diagnostic model. The results demonstrated that the MCP model achieved a prediction accuracy of 98.59% using only three miRNAs (hsa-miR-1343-3p, hsa-miR-5100, and hsa-miR-6765-5p). Consequently, model complexity was substantially reduced, and the model's generalization capability was improved. Biological validation revealed that these miRNAs were consistently selected across multiple models, furthermore, they are directly implicated in key pathways of gastric carcinogenesis, including the regulation of cell proliferation and apoptosis. This study provides a high-accuracy, cost-effective diagnostic strategy for early gastric cancer detection and identifies potential therapeutic targets.

Huiling Yang. Prediction of Early Gastric Cancer Based on miRNAs Using Penalized Logistic Regression. Academic Journal of Mathematical Sciences (2025), Vol. 6, Issue 2: 69-77. https://doi.org/10.25236/AJMS.2025.060210.

[1] Shen, J., Liao, J., Guarnera, M. A., Fang, H., Cai, L., Stass, S. A., & Jiang, F.•Analysis of MicroRNAs in Sputum to Improve Computed Tomography for Lung Cancer Diagnosis[J].Journal of Thoracic Oncology, 2014,9(1):33-40.

[2] Zaporozhchenko, I. A., Bryzgunova, O. E., Konoshenko, M. Y., Lekchnov, E. A., Amelina, E. V., Pashkovskaya, O. A., Yarmoschuk, S. V., Zheravin, A. A., Rykova, E. Y., & Laktionov, P. P.•Urine cell-free and extracellular vesicle cargo miRNAs as biomarkers for prostate cancer diagnosis[J]. Annals of Oncology, 2019,30(S5):v22.

[3] Koopaie, M., Manifar, S., Talebi, M. M., Kolahdooz, S., Razavi, A. E., Davoudi, M., & Pourshahidi, S.•Assessment of salivary miRNA, clinical, and demographic characterization in colorectal cancer diagnosis[J]. Translational Oncology, 2024,41:101880.

[4] Xi, C., Zhibin, H., Wenjing, W., Yi, B., Lijia, M., Chunni, Z., Cheng, W., Zhiji, R., Yang, Z., Sijia, W., Rui, Z., Yixin, Z., Heng, H., Chazhen, L., Lin, X., Jun, W., Hongbing, S., Junfeng, Z., Ke, Z., & Chen-Yu, Z.•Identification of ten serum microRNAs from a genome-wide serum microRNA expression profile as novel noninvasive biomarkers for nonsmall cell lung cancer diagnosis[J].•International journal of cancer, 2012, 130(7):1620-1628.

[5] Roberg, S. L. A., Damien, L., Andrea, Z., Mathias, W., Jonathan, W., & Charlotte, B.•A signature of miRNAs in the blood to help prognosticate prostate cancer at the time of diagnosis[J].•Journal of Clinical Oncology, 2017, 35(15):e16558.

[6] Chen, S., Lu, C., Lin, S., Sun, C., Wen, Z., Ge, Z., Chen, W., Li, Y., Zhang, P., Wu, Y., Wang, W., Zhou, H., Li, X., Lai, Y., & Li, H.•A panel based on three-miRNAs as diagnostic biomarker for prostate cancer[J]. Frontiers in Genetics, 2024, 15:1371441.

[7] Wang, Y., Chen, X., Jiang, W., Li, L., Li, W., Yang, L., Liao, M., Lian, B., Lv, Y., Wang, S., Wang, S., & Li, X.•Predicting human microRNA precursors based on an optimized feature subset generated by GA–SVM[J].•Genomics,•2011,•98(2):73-78.

[8] Bo, Z., Zhenmei, C., Baorui, T., Chenhe, Y., Zhifei, L., Yitong, L., Weiqing, S., Jing, L., & Jinhong, C.•m6A target microRNAs in serum for cancer detection[J].•Molecular Cancer,•2021,•20(1):170.

[9] Shen, L., & Tan, E. C.•Dimension reduction-based penalized logistic regression for cancer classification using microarray data[J].•Ieee-Acm Transactions on Computational Biology and Bioinformatics,2005,2(2):166-175.

[10] Liang, Y., Liu, C., Luan, X.-Z., Leung, K.-S., Chan, T.-M., Xu, Z.-B., & Zhang, H.•Sparse logistic regression with a L1/2 penalty for gene selection in cancer classification[J].•Bmc Bioinformatics, 2013,14

[11] Algamal, Z. Y., & Lee, M. H.•Regularized logistic regression with adjusted adaptive elastic net for gene selection in high dimensional cancer classification[J].•Computers in Biology and Medicine, 2015,•67:136-145.

[12] Lavanya, K., Rambabu, P., Suresh, G. V., & Bhandari, R.•Gene expression data classification with robust sparse logistic regression using fused regularisation[J].•International Journal of Ad Hoc and Ubiquitous Computing,•2023,•42(4):281-291.

[13] Breheny, P., & Huang, J.•Coordinate descent algorithms for nonconvex penalized regression, with applications to biological feature selection[J].The annals of applied statistics,•2011,•5(1):232.

[14] Wang, X., Li, Z., & Zhang, C.Integrated Analysis of Serum and Tissue microRNA Transcriptome for Biomarker Discovery in Gastric Cancer[J].Environmental toxicology,•2024,•)

[15] Zhang, Y., Wang, X., Liu, W., Lei, T., Qiao, T., Feng, W., & Song, W.CircGLIS3 promotes gastric cancer progression by regulating the miR-1343-3p/PGK1 pathway and inhibiting vimentin phosphorylation[J].Journal of Translational Medicine,2024,22(1):251.

[16] Zhang, H., Wang, J., Wang, Y., Li, J., Zhao, L., Zhang, T., & Liao, X.•Long non-coding LEF1-AS1 sponge miR-5100 regulates apoptosis and autophagy in gastric cancer cells via the miR-5100/DEK/AMPK-mTOR axis[J].•International Journal of Molecular Sciences,2022,23(9):4787.