Erişkin T Hücreli Lösemi (ATL) Patogenezindeki Marker Genlerin Makine Öğrenmesi Modelleri ile Keşfi ve Performans Karşılaştırması

Yıl 2025, Cilt: 15 Sayı: 3, 1046 - 1069, 15.09.2025

Sabire Kiliçarslan , Sait Can Yücebaş

https://doi.org/10.31466/kfbd.1597865

Öz

Hematolojik kanserler genellikle semptomlar belirginleştikten sonra teşhis edilir ve bu durum hastalığın kontrol altına alınmasını ve etkili tedavi stratejilerinin uygulanmasını zorlaştırabilir. Özellikle T hücreli lösemi gibi hematolojik kanserlerde, gen ekspresyon profillerinin incelenmesi, erken tanı ve tedavi stratejilerinin geliştirilmesinde hayati öneme sahiptir. Bu çalışma, Yetişkin T hücreli Lösemi (ATL) hücrelerinde ve sağlıklı bireylerin CD4+T hücrelerindeki tüm gen ekspresyon profilini karşılaştırarak, bu hastalığın patogenezindeki moleküler mekanizmaları farklı makine öğrenme yöntemleri ile ortaya çıkarma motivasyonu ile gerçekleştirilmiştir. Naive Bayes, K-En Yakın Komşu, Destek Vektör Makinesi, Rassal Orman, C4.5, Lojistik Regresyon, Doğrusal Diskriminant Analizi ve Yapay Sinir Ağları algoritmalarının karar performansları, GSE33615 veri seti üzerinde tabakalı örnekleme ile 5 katlı çapraz doğrulama yöntemi kullanılarak karşılaştırılmıştır. Bunlar arasında Yapay Sinir Ağı 0,98 AUC ve 0,93 F1 skoru ile öne çıkmıştır. Onu, 0.97 AUC ve 0.957 F1 skoru ile SVM takip etmiştir. Performans karşılaştırmasına ek olarak, ATL'ye neden olan genlerin tespiti için bilgi kazanç oranı, SHAPLEY metriği ve korelasyon değerleri hesaplanmıştır. Her model için en yüksek öneme sahip ilk on gen belirlenmiştir. Modeller tarafından önerilen genlerin kesişim kümesi dikkate alındığında, ZSCAN18, PLK3 ve NELL2 genlerinin ilgili hastalık için ilişkili olduğu bulunmuştur. Bu genler, hücre döngüsü düzenlenmesi, transkripsiyonel kontrol ve onkojenik sinyal iletimi üzerindeki rollerine bağlı olarak Erişkin T-hücreli Lösemi patogenezine katkıda bulunabilir. Bu genlerin moleküler rollerinin daha iyi anlaşılabilmesi için ileri araştırmalara ihtiyaç duyulmaktadır.

Anahtar Kelimeler

Yetişkin T-hücreli Lösemi (ATL) , Mikroarray çalışması , Makine öğrenmesi , Değişken önemi

Discovery of Marker Genes in Adult T Cell Leukemia (ATL) Pathogenesis with Machine Learning Models and Performance Comparison

Öz

Hematologic cancers are often diagnosed after symptoms become apparent, which can make it difficult to control the disease and implement effective treatment strategies. Studying gene expression profiles is vital for early diagnosis and the development of treatment strategies for hematologic cancers such as T-cell leukemia. The motivation of this study is to reveal the molecular mechanisms in the pathogenesis of this disease by comparing the whole gene expression profile in Adult T-cell Leukemia (ATL) cells and CD4+T cells of healthy individuals. For this aim, several machine learning algorithms, Naive Bayes, K-Nearest Neighbor, Support Vector Machine, Random Forest, C4.5, Logistic Regression, Linear Discriminant Analysis and Artificial Neural Network algorithms were used. Their performance was compared on the GSE33615 dataset by using 5-fold cross validation with stratified sampling. Among these, Artificial Neural Network stood out with an AUC of 0.98 and an F1 score of 0.93. It was followed by SVM with an AUC of 0.97 and 0.957 F1 score. In addition to performance comparison, information gain ratio, SHAPLEY metric and correlation values were calculated for the detection of genes causing ATL. Among the models, the three with the highest performance (ANN, SVM, RF) were selected, and the top ten most significant genes were identified for each. Considering the intersection of these gene sets, ZSCAN18, PLK3, and NELL2 were found to be associated with the related disease. These genes may contribute to Adult T-cell Leukemia pathogenesis through their roles in cell cycle regulation, transcriptional control, and oncogenic signaling. Further investigation is needed to clarify their precise molecular mechanisms in the related disease.

Anahtar Kelimeler

Adult T-cell Leukemia (ATL) , Microarray study , Machine learning , Variable importance

Kaynakça

• Abass, Y. A., & Adeshina, S. A. (2021). Deep learning methodologies for genomic data prediction. Journal of Artificial Intelligence for Medical Sciences, 2(1), 1-11
• Akalın, F., and Yumuşak, N. (2023). Mikrodizi veri kümesindeki ALL, AML ve MLL lösemi türlerine ilişkin gen anomalilerinin LSTM sinir ağı ile sınıflandırılması. Gazi Üniversitesi Mühendislik Mimarlık Fakültesi Dergisi, 38(3), 1299–1306.
• Breiman, L. (2001). Random Forests. Mach Learn, 45 (1): 5–32.
• Chi, C. M., Vossler, P., Fan, Y., & Lv, J. (2022). Asymptotic properties of high-dimensional random forests. The Annals of Statistics, 50(6), 3415-3438.
• Choi, H., Song, H., and Jung, Y. W. (2020). The roles of CCR7 for the homing of memory CD8+ T cells into their survival niches. Immune Network, 20(3).
• Chong, Y., Lee, J. Y., Kim, Y., Choi, J., Yu, H., Park, G., Cho, M. Y., and Thakur, N. (2020). A machine-learning expert-supporting system for diagnosis prediction of lymphoid neoplasms using a probabilistic decision-tree algorithm and immunohistochemistry profile database. Journal of Pathology and Translational Medicine, 54(6), 462–470.
• Cook, L., Rowan, A., & Bangham, C. (2021). ATLleukemia/lymphoma—Pathobiology and implications for modern clinical management. Annals of Lymphoma, 5.
• Cordo, V., Meijer, M. T., Hagelaar, R., de Goeij-de Haas, R. R., Poort, V. M., Henneman, A. A., Piersma, S. R., Pham, T. V., Oshima, K., and Ferrando, A. A. (2022). Phosphoproteomic profiling of T cell acute lymphoblastic leukemia reveals targetable kinases and combination treatment strategies. Nature Communications, 13(1), 1048.
• Cover, T., & Hart, P. (1967). Nearest neighbor pattern classification. IEEE transactions on information theory, 13(1), 21-27.
• Eckardt, J. N., Bornhäuser, M., Wendt, K., and Middeke, J. M. (2020). Application of machine learning in the management of acute myeloid leukemia: current practice and future prospects. Blood Advances, 4(23), 6077-6085.
• Ekanayake, I. U., Meddage, D. P. P., and Rathnayake, U. (2022). A novel approach to explain the black-box nature of machine learning in compressive strength predictions of concrete using Shapley additive explanations (SHAP). Case Studies in Construction Materials, 16, e01059.
• Erdem, E., & Bozkurt, F. (2021). A comparison of various supervised machine learning techniques for prostate cancer prediction. Avrupa Bilim ve Teknoloji Dergisi, (21), 610-620.
• Fahim, N. I., Utsha, M. A. H., Karmaker, R. S., Ullah, M. O., and Farid, D. M. (2023). Decision Tree using Feature Grouping. 2023 26th International Conference on Computer and Information Technology (ICCIT) (pp. 1–5). Cox's Bazar, Bangladesh.
• Faiz, M., Mounika, B. G., Akbar, M., and Srivastava, S. (2024). Deep and Machine Learning for Acute Lymphoblastic Leukemia Diagnosis: A Comprehensive Review. Advances in Distributed Computing and Artificial Intelligence Journal, 13, e31420-e31420.
• Fisher, R. A. (1936). The use of multiple measurements in taxonomic problems. Annals of eugenics, 7(2), 179-188.
• Frank, D. A. (1999). STAT Signaling in the Pathogenesis and Treatment of Cancer. Molecular Medicine, 5(7), 432–456. https://doi.org/10.1007/BF03403538
• Fujikawa, D., Nakagawa, S., Hori, M., Kurokawa, N., Soejima, A., Nakano, K., Yamochi, T., Nakashima, M., Kobayashi, S., and Tanaka, Y. (2016). Polycomb-dependent epigenetic landscape in ATLleukemia. Blood, The Journal of the American Society of Hematology, 127(14), 1790–1802.
• Gessain, A., and Cassar, O. (2012). Epidemiological aspects and world distribution of HTLV-1 infection. Frontiers in Microbiology, 3, 388.
• Ghobadi, M. Z., Emamzadeh, R., and Afsaneh, E. (2022). Exploration of mRNAs and miRNA classifiers for various ATLL cancer subtypes using machine learning. BMC Cancer, 22(1), 433. https://doi.org/10.1186/s12885-022-09540-1
• Goroshchuk, O., Kolosenko, I., Vidarsdottir, L., Azimi, A., & Palm-Apergi, C. (2019). Polo-like kinases and acute leukemia. Oncogene, 38(1), 1-16.)
• Guido, R., Ferrisi, S., Lofaro, D., and Conforti, D. (2024). An Overview on the Advancements of Support Vector Machine Models in Healthcare Applications: A Review. Information, 15(4), 235.
• Guo, W., Liu, R., Ono, Y., Ma, A.-H., Martinez, A., Sanchez, E., Wang, Y., Huang, W., Mazloom, A., and Li, J. (2012). Molecular characteristics of CTA056, a novel interleukin-2-inducible T-cell kinase inhibitor that selectively targets malignant T cells and modulates oncomirs. Molecular Pharmacology, 82(5), 938–947.
• Goodfellow, I., Bengio, Y., Courville, A., & Bengio, Y. (2016). Deep learning (Vol. 1, No. 2). Cambridge: MIT press.
• Hall, E. S. (2021). Applying Polygenic Models to Disentangle Genotype-Phenotype Associations across Common Human Diseases. (unpublished master’sdissertation). University of Toronto, Canada
• Haury, A. C., Gestraud, P., & Vert, J. P. (2011). The influence of feature selection methods on accuracy, stability and interpretability of molecular signatures. PloS one, 6(12), e28210.
• Haykin, S. (1994). Neural networks: a comprehensive foundation. Prentice Hall PTR
• Helmke, C., Becker, S., and Strebhardt, K. (2016). The role of Plk3 in oncogenesis. Oncogene, 35(2), 135–147.
• Hosmer Jr, D. W., Lemeshow, S., & Sturdivant, R. X. (2013). Applied logistic regression. John Wiley & Sons.
• Hukasova, E. (2017). Cell Cycle Regulation and DNA Damage Response: A Record of Polo-Like Kinase 1 Activity.(unpublished doctoral dissertation). Karolinska Institutet, Stockholm, Sweden
• Hyun, T. S., and Ross, T. S. (2004). HIP1: Trafficking roles and regulation of tumorigenesis. Trends in Molecular Medicine, 10(4), 194–199.
• Inan, M. S. K., and Rahman, I. (2023). Explainable AI integrated feature selection for landslide susceptibility mapping using TreeSHAP. SN Computer Science, 4(5), 482.
• Ishitsuka, K., and Tamura, K. (2014). Human T-cell leukaemia virus type I and ATLleukaemia-lymphoma. The Lancet Oncology, 15(11), e517–e526.
• Jayabal, P., Zhou, F., Lei, X., Ma, X., Blackman, B., Weintraub, S. T., ... & Shiio, Y. (2021). NELL2-cdc42 signaling regulates BAF complexes and Ewing sarcoma cell growth. Cell reports, 36(1)
• Katsuya, H., Ishitsuka, K., Utsunomiya, A., Hanada, S., Eto, T., Moriuchi, Y., Saburi, Y., Miyahara, M., Sueoka, E., and Uike, N. (2015). Treatment and survival among 1594 patients with ATL. Blood, The Journal of the American Society of Hematology, 126(24), 2570–2577.
• Keller, J. M., Gray, M. R., & Givens, J. A. (1985). A fuzzy k-nearest neighbor algorithm. IEEE transactions on systems, man, and cybernetics, (4), 580-585.
• Kılıçarslan, S., and Pacal, I. (2023). Domates Yapraklarında Hastalık Tespiti İçin Transfer Öğrenme Metotlarının Kullanılması. Mühendislik Bilimleri ve Araştırmaları Dergisi, 5(2), 215–222.
• Kim, D. H., Roh, Y.-G., Lee, H. H., Lee, S.-Y., Kim, S. I., Lee, B. J., and Leem, S.-H. (2013). The E2F1 Oncogene Transcriptionally Regulates NELL2 in Cancer Cells. DNA and Cell Biology, 32(9), 517–523. https://doi.org/10.1089/dna.2013.1974
• Kurosawa, N., Fujimoto, R., Ozawa, T., Itoyama, T., Sadamori, N., & Isobe, M. (2013). Reduced level of the BCL11B protein is associated with adult T-cell leukemia/lymphoma. PLoS One, 8(1), e55147.)
• LeBlanc, F. R., Pearson, J. M., Tan, S. F., Cheon, H., Xing, J. C., Dunton, W., ... & Loughran Jr, T. P. (2020). Sphingosine kinase‐2 is overexpressed in large granular lymphocyte leukaemia and promotes survival through Mcl‐1. British journal of haematology, 190(3), 405-417.) .
• LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. nature, 521(7553), 436-444.
• Lee, H. W., Lawton, C., Na, Y. J., & Yoon, S. (2013). Robustness of chemometrics-based feature selection methods in early cancer detection and biomarker discovery. Statistical applications in genetics and molecular biology, 12(2), 207-223
• Lee, S., Kim, J., Kang, H., Kang, D. Y., & Park, J. (2021). Genetic algorithm based deep learning neural network structure and hyperparameter optimization. Applied Sciences, 11(2), 744.
• Legler, D. F., Uetz-von Allmen, E., and Hauser, M. A. (2014). CCR7: Roles in cancer cell dissemination, migration and metastasis formation. The International Journal of Biochemistry and Cell Biology, 54, 78–82.
• Li, B., Ren, B., Ma, G., Cai, F., Wang, P., Zeng, Y., ... & Deng, J. (2023). Inactivation of ZSCAN18 by promoter hypermethylation drives the proliferation via attenuating TP53INP2-mediated autophagy in gastric cancer cells. Clinical epigenetics, 15(1), 10.)
• Li, C., Zhao, J., Kang, B., Li, S., Tang, J., Dong, D., & Chen, Y. (2024). Identification and validation of STAT4 as a prognostic biomarker in acute myeloid leukemia. Bioscience Reports, 44(2).
• Liu, G., Li, J., Zhang, C. Y., Huang, D. Y., & Xu, J. W. (2021). ARHGAP20 expression inhibited HCC progression by regulating the PI3K-akt signaling pathway. Journal of Hepatocellular Carcinoma, 271-284
• Liu, Y., Wang, X., Deng, L., Ping, L., Shi, Y., Zheng, W., Lin, N., Wang, X., Tu, M., Xie, Y., Liu, W., Ying, Z., Zhang, C., Pan, Z., Wang, X., Ding, N., Song, Y., and Zhu, J. (2019). ITK inhibition induced in vitro and in vivo anti-tumor activity through downregulating TCR signaling pathway in malignant T cell lymphoma. Cancer Cell International, 19(1), 32. https://doi.org/10.1186/s12935-019-0754-9
• Lundberg, S. (2017). A unified approach to interpreting model predictions. arXiv Preprint arXiv:1705.07874.
• Lundberg, S. M., Erion, G. G., and Lee, S.-I. (2019). Consistent Individualized Feature Attribution for Tree Ensembles. arXiv Preprint arXiv. http://arxiv.org/abs/1802.03888
• Miao, M., Li, S., Yu, Y., and Li, F. (2023). LysM-containing proteins function in the resistance of Litopenaeus vannamei against Vibrio parahaemolyticus infection. Developmental and Comparative Immunology, asa
• Nakamura, R., Oyama, T., Tajiri, R., Mizokami, A., Namiki, M., Nakamoto, M., and Ooi, A. (2015). Expression and regulatory effects on cancer cell behavior of NELL1 and NELL2 in human renal cell carcinoma. Cancer Science, 106(5), 656–664. https://doi.org/10.1111/cas.12649
• Patel, S., Patel, H., Vyas, D., and Degadwala, S. (2021). Multi-Classifier Analysis of Leukemia Gene Expression From Curated Microarray Database (CuMiDa). 2021 2nd International Conference on Smart Electronics and Communication (ICOSEC), (pp. 1174–1178). Trichy, India
• Quinlan, J. R. (2014). C4. 5: Programs for machine learning. Elsevier.
• Rajasingh, J., Raikwar, H. P., Muthian, G., Johnson, C., and Bright, J. J. (2006). Curcumin induces growth-arrest and apoptosis in association with the inhibition of constitutively active JAK–STAT pathway in T cell leukemia. Biochemical and Biophysical Research Communications, 340(2), 359–368.
• Roy, A., and Chakraborty, S. (2023). Support vector machine in structural reliability analysis: A review. Reliability Engineering and System Safety, 233, 109126.
• Rish, I. (2001, August). An empirical study of the naive Bayes classifier. In IJCAI 2001 workshop on empirical methods in artificial intelligence (Vol. 3, No. 22, pp. 41-46).
• Rupapara, V., Rustam, F., Aljedaani, W., Shahzad, H. F., Lee, E., & Ashraf, I. (2022). Blood cancer prediction using leukemia microarray gene data and hybrid logistic vector trees model. Scientific reports, 12(1), 1000. https://doi.org/10.1038/s41598-022-04835-6
• Sakamoto, Y., Ishida, T., Masaki, A., Murase, T., Ohtsuka, E., Takeshita, M., ... & Inagaki, H. (2022). CCR7 alterations associated with inferior outcome of adult T‐cell leukemia/lymphoma under mogamulizumab treatment. Hematological Oncology, 40(5), 876-884
• Saralamma, V. V., Vetrivel, P., Lee, H., Kim, S., Ha, S., Murugesan, R., Kim, E., Heo, J., and Kim, G. (2020). Comparative proteomic analysis uncovers potential biomarkers involved in the anticancer effect of Scutellarein in human gastric cancer cells. Oncology Reports, 44(3), 939–958. https://doi.org/10.3892/or.2020.7677.
• Stagno, F., Russo, S., Murdaca, G., Mirabile, G., Alvaro, M. E., Nasso, M. E., ... & Allegra, A. (2025). Utilization of Machine Learning in the Prediction, Diagnosis, Prognosis, and Management of Chronic Myeloid Leukemia. International Journal of Molecular Sciences, 26(6), 2535.
• Stricker, S. H., Köferle, A., & Beck, S. (2017). From profiles to function in epigenomics. Nature Reviews Genetics, 18(1), 51-66.
• Sundaramurthi, H., Tonelotto, V., Wynne, K., O’Connell, F., O’Reilly, E., Costa-Garcia, M., Kovácsházi, C., Kittel, A., Marcone, S., and Blanco, A. (2023). Ergolide mediates anti-cancer effects on metastatic uveal melanoma cells and modulates their cellular and extracellular vesicle proteomes. Open Research Europe, 3.
• Takasaki, Y., Iwanaga, M., Imaizumi, Y., Tawara, M., Joh, T., Kohno, T., Yamada, Y., Kamihira, S., Ikeda, S., and Miyazaki, Y. (2010). Long-term study of indolent ATLleukemia-lymphoma. Blood, The Journal of the American Society of Hematology, 115(22), 4337–4343.
• Uchiyama, T., Yodoi, J., Sagawa, K., Takatsuki, K., and Uchino, H. (1977). ATLleukemia: Clinical and hematologic features of 16 cases. Blood, 50(3), 481–492.
• Vapnik, V. (1995). Support-vector networks. Machine Learning, 20, 273–297.
• Wang, Y., Luo, Y., Fu, S., He, L., Pan, G., Fan, D., Wen, Q., and Fan, Y. (2023). Zinc finger and SCAN domain-containing protein 18 is a potential DNA methylation-modified tumor suppressor and biomarker in breast cancer. Frontiers in Endocrinology, 14, 1095604.
• Wang, H., Liang, Q., Hancock, J. T., & Khoshgoftaar, T. M. (2024). Feature selection strategies: a comparative analysis of SHAP-value and importance-based methods. Journal of Big Data, 11(1), 44.
• Watanabe, M., Nakahata, S., Hamasaki, M., Saito, Y., Kawano, Y., Hidaka, T., ... & Morishita, K. (2010). Downregulation of CDKN1A in adult T-cell leukemia/lymphoma despite overexpression of CDKN1A in human T-lymphotropic virus 1-infected cell lines. Journal of virology, 84(14), 6966-6977.
• Xie, S., Wu, H., Wang, Q., Cogswell, J. P., Husain, I., Conn, C., Stambrook, P., Jhanwar-Uniyal, M., and Dai, W. (2001). Plk3 functionally links DNA damage to cell cycle arrest and apoptosis at least in part via the p53 pathway. Journal of Biological Chemistry, 276(46), 43305–43312.
• Xu, H., Jia, J., Jeong, H.-H., and Zhao, Z. (2023). Deep learning for detecting and elucidating human T-cell leukemia virus type 1 integration in the human genome. Patterns, 4(2).
• Yamagishi, M., Nakano, K., Miyake, A., Yamochi, T., Kagami, Y., Tsutsumi, A., Matsuda, Y., Sato-Otsubo, A., Muto, S., and Utsunomiya, A. (2012). Polycomb-mediated loss of miR-31 activates NIK-dependent NF-κB pathway in adult T cell leukemia and other cancers. Cancer Cell, 21(1), 121–135.
• Zhang, L., Zhou, L., Wang, Y., Li, C., Liao, P., Zhong, L., ... & Weng, J. (2022). Deep learning-based transcriptome model predicts survival of T-cell acute lymphoblastic leukemia. Frontiers in Oncology, 12, 1057153
• Zhang, X., Wei, C., Liang, H., & Han, L. (2021). Polo-like kinase 4’s critical role in cancer development and strategies for Plk4-targeted therapy. Frontiers in Oncology, 11, 587554.
• Zhao, X., Lu, M., Liu, Z., Zhang, M., Yuan, H., Dan, Z., ... & Dang, C. (2023). Comprehensive analysis of alfa defensin expression and prognosis in human colorectal cancer. Frontiers in oncology, 12, 974654.).
• Zhu, X. L., Zeng, Y. F., Guan, J., Li, Y. F., Deng, Y. J., Bian, X. W., ... & Liang, L. (2011). FMNL2 is a positive regulator of cell motility and metastasis in colorectal carcinoma. The Journal of pathology, 224(3), 377-388