Nanoparticle-induced toxicity in pancreatic cancer: Insights from machine learning models

Date

2025-12-01

Author

Cakir, Irem
Yildirim, Elif
İleri Ercan, Nazar

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

16
views

0
downloads

Cite This

In this study, 83 research articles published between 2008 and 2024 on the effects of various nanoparticles on pancreatic cancer are reviewed, and a dataset of 2696 data points is constructed. The dataset, which is analyzed using Machine Learning (ML) models, includes the effects of 20 different nanoparticles on cell viability and apoptosis in 29 different pancreatic cell lines. After feature selection with the Boruta algorithm, a Random Forest model with 19 features is trained to predict cell viability. The feature selection results indicate that drug type, drug concentration, nanoparticle concentration, and nanoparticle material are the most significant predictors of cytotoxicity. Additionally, Association Rule Mining (ARM) reveals that BSA nanoparticles are linked to cytotoxic outcomes. Further double-factor ARM analysis shows a strong correlation between BSA as a core material and BxPC-3 cells with cytotoxicity, while gold nanoparticles at lower concentrations (<1000 μM) are strongly associated with non-cytotoxicity. Moreover, high drug and nanoparticle concentrations exceeding 1000 μM are consistently linked to cell viabilities below 40 %, highlighting their potent cytotoxic effects. These findings provide valuable insights into the key factors influencing cytotoxicity in pancreatic cells induced by nanoparticles, paving the way for more effective therapeutic strategies.

Subject Keywords

Machine learning, Nanoparticles, Pancreatic cancer

URI

https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=105018684182&origin=inward
https://hdl.handle.net/11511/116692

Collections

Department of Chemical Engineering, Article