This study investigates the use of Neural Ordinary Differential Equations (NODEs) as an alternative to traditional compartmental models and Nonlinear Mixed-Effects (NLME) models for drug concentration prediction in pharmacokinetics. Unlike standard models that rely on strong assumptions and often struggle with high-dimensional covariate relationships, NODEs offer a data-driven approach, learning differential equations directly from data while integrating covariates. To evaluate their performance, NODEs were applied to a real-world Dalbavancin pharmacokinetic dataset comprising 218 patients and compared against a two-compartment model and an NLME within a cross-validation framework, which ensures an evaluation of robustness. Given the challenge of limited data availability, a data augmentation strategy was employed to pre-train NODEs. Their predictive performance was assessed both with and without covariates, while model explainability was analyzed using Shapley additive explanations (SHAP) values. Results show that, in the absence of covariates, NODEs performed comparably to state-of-the-art NLME models. However, when covariates were incorporated, NODEs demonstrated superior predictive accuracy. SHAP analyses further revealed how NODEs leverage covariates in their predictions. These results establish NODEs as a promising alternative for pharmacokinetic modeling, particularly in capturing complex covariate interactions, even when dealing with sparse and small datasets, thus paving the way for improved drug concentration predictions and personalized treatment strategies in precision medicine.

Hematological malignancies are a diverse group of cancers developing in the peripheral blood, the bone marrow or the lymphatic system. Due to their heterogeneity, the identification of novel and advanced molecular signatures is essential for enhancing their characterization and facilitate its translation to new pharmaceutical solutions and eventually to clinical applications. In this study, we collected publicly available microarray data for more than five thousand subjects, across thirteen hematological malignancies. Using PANDA to estimate gene regulatory networks (GRNs), we performed hierarchical clustering and network analysis to explore transcription factor (TF) interactions and their implications on biological pathways. Our findings reveal distinct clustering patterns among leukemias and lymphomas, with notable differences in gene and TF expression profiles. Gene Set Enrichment Analysis (GSEA) identified 57 significantly enriched KEGG pathways, highlighting both common and unique biological processes across HMs. We also identified potential drug targets within these pathways, emphasizing the role of TFs such as CEBPB and NFE2L1 in disease pathophysiology. Our comprehensive analysis enhances the understanding of the molecular landscape of HMs and suggests new avenues for targeted therapeutic strategies. These findings also motivate the adoption of regulatory networks, combined with modern biotechnological possibilities, for insightful pan-cancer exploratory studies.