Nazanin Ahmadi Daryakenari

Nazanin Ahmadi Daryakenari

Auto-generated profile, based on publicly available information

Nazanin Ahmadi Daryakenari

Center for Biomedical Engineering, School of Engineering, Brown University , Providence, RI

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
12 March 2024 in PLOS Computational Biology
Verified
Journal Article
Open access
PDF file
NDNazanin Ahmadi Daryakenari
Nazanin Ahmadi Daryakenari
Mario De FlorioMario De Florio
Mario De Florio
KSKhemraj Shukla
Khemraj Shukla
George KarniadakisGeorge Karniadakis
George Karniadakis
Discovering mathematical equations that govern physical and biological systems from observed data is a fundamental challenge in scientific research. We present a new physics-informed framework for parameter estimation and missing physics identification (gray-box) in the field of Systems Biology. The proposed framework—named AI-Aristotle—combines the eXtreme Theory of Functional Connections (X-TFC) domain-decomposition and Physics-Informed Neural Networks (PINNs) with symbolic regression (SR) techniques for parameter discovery and gray-box identification. We test the accuracy, speed, flexibility, and robustness of AI-Aristotle based on two benchmark problems in Systems Biology: a pharmacokinetics drug absorption model and an ultradian endocrine model for glucose-insulin interactions. We compare the two machine learning methods (X-TFC and PINNs), and moreover, we employ two different symbolic regression techniques to cross-verify our results. To test the performance of AI-Aristotle, we use sparse synthetic data perturbed by uniformly distributed noise. More broadly, our work provides insights into the accuracy, cost, scalability, and robustness of integrating neural networks with symbolic regressors, offering a comprehensive guide for researchers tackling gray-box identification challenges in complex dynamical systems in biomedicine and beyond.

28 November 2024 in Computers in Biology and Medicine
Verified
Journal Article
Closed access
NDNazanin Ahmadi Daryakenari
Nazanin Ahmadi Daryakenari
SWShupeng Wang
Shupeng Wang
George KarniadakisGeorge Karniadakis
George Karniadakis

19 December 2025 in Neural Networks
Verified
Journal Article
Closed access
ZHZheyuan Hu
Zheyuan Hu
NDNazanin Ahmadi Daryakenari
Nazanin Ahmadi Daryakenari
QSQianli Shen
Qianli Shen
KKKenji Kawaguchi
Kenji Kawaguchi
George KarniadakisGeorge Karniadakis
George Karniadakis

24 December 2025 in Computers in Biology and Medicine
Verified
Journal Article
Open access
NDNazanin Ahmadi Daryakenari
Nazanin Ahmadi Daryakenari
KSKhemraj Shukla
Khemraj Shukla
George KarniadakisGeorge Karniadakis
George Karniadakis
Physics-Informed Kolmogorov–Arnold Networks (PIKANs) are gaining attention as an effective counterpart to the original multilayer perceptron-based Physics-Informed Neural Networks (PINNs). Both representation models can address inverse problems and facilitate gray-box system identification. However, a comprehensive understanding of their performance in terms of accuracy and speed remains underexplored. In particular, we introduce a modified PIKAN architecture, tanh-cPIKAN, which is based on Chebyshev polynomials for parametrization of the univariate functions with an extra nonlinearity for enhanced performance. We then present a systematic investigation of how choices of the optimizer, representation, and training configuration influence the performance of PINNs and PIKANs in the context of systems pharmacology modeling. We benchmark a wide range of first-order, second-order, and hybrid optimizers—including various learning rate schedulers. We use the new Optax library [1] to identify the most effective combinations for learning gray-boxes under ill-posed, non-unique, and data-sparse conditions. We examine the influence of model architecture (MLP vs. KAN), numerical precision (single vs. double), the need for warm-up phases for second-order methods, and sensitivity to the initial learning rate. We also assess the optimizer scalability for larger models and analyze the trade-offs introduced by JAX in terms of computational efficiency and numerical accuracy. Using two representative systems pharmacology case studies — a pharmacokinetics model and a chemotherapy drug-response model — we offer practical guidance on selecting optimizers and representation models/architectures for robust and efficient gray-box discovery. Our findings provide actionable insights for improving the training of physics-informed networks in biomedical applications and beyond.

24 June 2026 in Computer Methods in Applied Mechanics and Engineering
Verified
Journal Article
Open access
SKSiavash Khodakarami
Siavash Khodakarami
VOVivek Oommen
Vivek Oommen
NDNazanin Ahmadi Daryakenari
Nazanin Ahmadi Daryakenari
MBMaxim Beekenkamp
Maxim Beekenkamp
GKGeorge Em Karniadakis
George Em Karniadakis

30 September 2026 in PNAS Nexus
Verified
Journal Article
Open access
ZCZhaojie Chai
Zhaojie Chai
NDNazanin Ahmadi Daryakenari
Nazanin Ahmadi Daryakenari
George KarniadakisGeorge Karniadakis
George Karniadakis
Red blood cell (RBC) clearance by macrophages maintains blood homeostasis and is dysregulated in the hemolytic disorder sickle cell disease (SCD) and the lysosomal storage disorder Gaucher disease (GD), where biophysical and biochemical alterations promote premature phagocytosis. We develop a multiscale hybrid modeling framework integrating signaling dynamics, biophysical simulations, and machine learning to investigate the mechanisms governing RBC phagocytosis in these diseases. Our approach couples a systems biology model of macrophage–RBC signaling with dissipative particle dynamics (DPD) simulations of molecular diffusion and membrane interactions, and leverages Physics-Informed Neural Networks (PINNs) for simultaneous parameter inference, hidden-state reconstruction, and integration of multiscale mechanistic constraints. The DPD framework provides mechanistic insight into antibody diffusion, receptor engagement, and membrane-level interactions during macrophage–RBC contact, generating spatially resolved trajectories of CD47–SIRPα signaling and antibody–receptor binding that serve as intermediate observables constraining the signaling model. The model accurately captures differential phagocytic responses between healthy and altered RBCs, revealing diminished inhibitory signaling and changes in SHP1-mediated pathways in both SCD and GD. Identifiability analysis combining Fisher Information Matrix diagnostics and profile likelihood confirms that parameters governing the CD47–SIRPα–SHP1 axis are among the most robustly recoverable, and simulations of therapeutic perturbations with anti-SIRPα antibodies demonstrate modulation of engulfment outcomes. We further employ Physics-Informed Kolmogorov-Arnold Networks (PIKANs) as an alternative to standard PINNs, demonstrating improved robustness under noise and sampling variability. More broadly, our multiscale platform linking biophysical simulation with systems-level inference is generalizable, offering mechanistic insights and computational tools for therapeutic exploration in diseases involving dysregulated phagocytosis.