High Expression of Caspase-8 Associated with Improved Survival in Diffuse Large B-Cell Lymphoma: Machine Learning and Artificial Neural Networks Analyses

• DOI: 10.3390/biomedinformatics1010003
• Publication Date: 2021-04-22T01:25:10Z
• Authors: Joaquim Carreras; Yara Yukie Kikuti; Giovanna Roncador; Masashi Miyaoka; Shinichiro Hiraiwa; Sakura Tomita; Haruka Ikoma; Yusuke Kondo; Atsushi Ito; Sawako Shiraiwa; Kiyoshi Ando; Naoya Nakamura; Rifat Hamoudi
• Keywords: Diffuse Large B-cell Lymphoma; apoptosis; diffuse large B-cell lymphoma; personalized and precision medicine; explainable Artificial Intelligence; artificial intelligence; 3. Good health; 03 medical and health sciences; machine learning; 0302 clinical medicine; Caspase-3; prognosis; Caspase-8; diffuse large B-cell Lymphoma; artificial neural network

High expression of the anti-apoptotic TNFAIP8 is associated with poor survival patients diffuse large B-cell lymphoma (DLBCL), and one functions to inhibit pro-apoptosis Caspase-8. We aimed analyze immunohistochemical Caspase-8 (active subunit p18; CASP8) in a series 97 cases DLBCL from Tokai University Hospital, correlate other pathway-related markers, including cleaved Caspase-3, PARP, BCL2, TP53, MDM2, MYC, Ki67, E2F1, CDK6, MYB LMO2. After digital image quantification, correlation several clinicopathological characteristics showed that high protein was favorable overall progression-free (Hazard Risks = 0.3; p 0.005 0.03, respectively). also positively correlated cCASP3, TNFAIP8, BCL2 Ki67. Next, modeled using predictive analytics, accuracy (>80%) obtained CHAID decision tree, Bayesian network, discriminant analysis, C5 logistic regression, Artificial Intelligence Neural Network methods (both Multilayer perceptron Radial basis function); most relevant markers were cPARP, TNFAIP8. Finally, CASP8 gene successfully an independent 414 Lymphoma/Leukemia Molecular Profiling Project (LLMPP). In conclusion, prognosis DLBCL. Predictive modeling feasible analytic strategy results solution can be understood (i.e., explainable artificial intelligence, “white-box” algorithms).