Page 29 - Research and Innovation Pulse

Mechanistic Insights into the Pharmacological Actions of

Flavonoids: A Comprehensive Review

Author(s):
ViT-Stain: Vision Transformer-Driven Virtual Staining for Skin
Rasheed, A. (University of Peshawar), Ali, G. (University of Peshawar), Islam, M. R. (Daffodil International University), Rauf, Histopathology via Global Contextual Learning
A. (University of Swabi), Ajaj, R. (Abu Dhabi University), Hemeg, H. A. (Taibah University), Iriti, M. (University of Milan /
National Interuniversity Consortium of Materials Science and Technology)

Index Terms: Author(s):

antidiabetic agent; antioxidant; flavonoid; glycosylated protein; liposome; mitogen activated protein kinase; nanoparticle; Hussain, M. A. (National University of Sciences and Technology), Waris, M. A. (National University of Sciences and
phosphatidylinositol 3 kinase; protein kinase B; transcription factor Nrf2; antineoplastic activity; apoptosis; autophagy Technology), Akram, M. U. (National University of Sciences and Technology), Khan, M. J. (National University of Sciences
(cellular); B lymphocyte; bioavailability; biological activity; cancer inhibition; cardiovascular disease; cell protection; chronic and Technology), Asaf, M. Z. (National University of Sciences and Technology), Javaid, A. (National University of Sciences
disease; degenerative disease; diabetes mellitus; drug delivery device; drug delivery system; drug development; drug and Technology), Gilani, S. O. (Abu Dhabi University), Hazzazi, F. (Prince Sattam Bin Abdul Aziz University)
mechanism; enhancer region; gene expression; glycosylation; human; hydroxylation; inflammation; molecular docking;
neuroprotection; nonhuman; oxidative stress; personalized medicine; pharmacokinetics; pharmacology; prevention; Index Terms:
prophylaxis; review; signal transduction; structure activity relation; systems pharmacology
adult; aged; area under the curve; Article; artificial intelligence; artificial neural network; basal cell carcinoma;
Abstract: benchmarking; controlled study; convolutional neural network; cost effectiveness analysis; deep learning; diagnostic test
accuracy study; entropy; geometry; global contextual learning; global health; hallucination; histology; histopathology;

Flavonoids, a diverse group of polyphenolic chemicals found in plants, have significant attention for their diverse pharmacological human; human tissue; image quality; image reconstruction; image segmentation; learning; learning algorithm;
actions and therapeutic potential. Their ability to target multiple pathways, modulate oxidative stress, and regulate inflammatory multilayer perceptron; photometry; qualitative research; receiver operating characteristic; receptive field; skin biopsy;
mediators is crucial in preventing and managing chronic diseases like cancer, cardiovascular disorders, diabetes, and neurodegenerative spatial analysis; squamous cell carcinoma; training; vision; image processing; pathology; procedures; skin; staining;
diseases. Flavonoids have multitargeted actions, providing a safer and general therapeutic approach compared to single-targeted hematoxylin; Hematoxylin; Humans; Image Processing, Computer-Assisted; Neural Networks, Computer; Skin; Staining
synthetic drugs. This review provides a comprehensive understanding of flavonoids’ biological effects, focusing on their modulation and Labeling
of key molecular signaling pathways such as nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), mitogen-activated
protein kinases (MAPK), phosphatidylinositol 3-kinase (PI3 K)/protein kinase B (AKT), nuclear factor erythroid 2-related factor 2 (Nrf2), Abstract:
oxidative stress, inflammation, and apoptosis. Their anticancer potential is supported by their ability to induce apoptosis, modulate
autophagy, and influence gene expression, while their anti-inflammatory and antioxidant properties aid in cytoprotection. Certain Current virtual staining approaches for histopathology slides use convolutional neural networks (CNNs) and generative
hydroxylation and glycosylation patterns enhance their biological efficacy based on structure-activity connections. The review adversarial networks (GANs). These approaches rely on local receptive fields, struggle to capture global context, and long-
demonstrates the various benefits of these substances, including their hepatoprotective, neuroprotective, anticancer, antidiabetic, range tissue dependencies. This limitation can introduce artifacts in fine textures and cause loss of subtle morphological
and cardioprotective properties, based on both experimental and clinical evidence. It discusses the structure–activity relationship details. We propose a novel vision transformer-driven virtual staining framework (ViT-Stain) that translates unstained
(SAR) that supports their bioefficacy as well as issues with metabolism, bioavailability, and therapeutic translation. It also provides skin tissue images into hematoxylin and eosin (H&E)-equivalent images. The transformer’s self-attention enables ViT-
a comprehensive understanding of flavonoids as potential agents for chronic disease prevention and management, integrating Stain to capture long-range dependencies, preserve global context, and maintain fine textures. We trained ViT-Stain on
pharmacological findings with molecular facts. A method was used to identify works published in reputable journals. Every search result the E-Staining DermaRepo dataset, which pairs unstaained and H&E-stained whole-slide images (WSIs). We validated
came from PubMed, Scopus, Web of Science, ScienceDirect, Google Scholar, etc The terms flavonoids, pharmacological properties, our model using metrics including SSIM, PSNR, FID, KID, LPIPS, and a novel histology-specific fidelity index (HSFI). Three
disease, and mechanism of action were utilized. We selected and investigated research papers, review articles, and original studies board-certified pathologists provided feedback for qualitative evaluations. ViT-Stain outperforms leading CNN and GAN
that were published up until 2025. Future research should focus on improving flavonoids’ bioavailability using advanced drug delivery models, including Pix2Pix, CycleGAN, CUTGAN, and DCLGAN. It achieves an overall diagnostic concordance of 85% with
methods like conjugates, liposomes, and nanoparticles, with extensive clinical trials needed for validation. Furthermore, the potential virtual H&E-stains (Fleiss’ κ=0.88). However, the model requires longer training (about 93 hours on A100 GPUs) and
of flavonoids in therapeutic interventions will be enhanced through the use of computational techniques such as molecular docking, inference times (about 2.9 minutes). Our work advances AI-driven diagnostic reproducibility for high-fidelity clinical
network pharmacology, and precision medicine. Future perspectives emphasize the need for advanced drug delivery systems, clinical settings and aligns with the World Health Organization (WHO) global health goals.
trials, and molecular docking techniques to enhance their therapeutic efficacy.
Read the paper
Read the paper

Abu Dhabi University | Research and Innovation Pulse Newsletter 29