To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Random Forest Modeling of Curcumin-Loaded PLGA Nanoparticle Size: Mechanistic Insights and In-Silico Formulation Screening |
|
Author |
Rumana Ferdushi |
| Citation |
Vol. 26 No. 2 pp. 17-22 |
|
Abstract |
The therapeutic potential of curcumin and the biodegradability of poly (lactic-co-glycolic acid) (PLGA) have led to extensive research on curcumin-loaded PLGA nanoparticles for drug delivery applications. The intricate relationships between formulation composition and processing factors make it difficult to precisely control nanoparticle size during nanoprecipitation. Using experimentally determined nanoprecipitation factors, Random Forest regression was used in this study to establish a data-driven modeling framework for predicting nanoparticle size. After duplicate aggregation, 19 distinct curcumin-loaded PLGA formulations were examined. Significantly surpassing linear regression (R2 ? 0.04), Leave-One-Out Cross-Validation (LOOCV) produced moderate predictive performance (R2 ? 0.32), suggesting nonlinear correlations between process factors and particle size. According to feature importance analysis, vortex duration had little effect within the studied range, while second centrifugation speed was the main contributor, followed by PLGA concentration. A partial dependence analysis showed non-monotonic behavior related to centrifugation speed and a monotonic size increase with polymer concentration. The experimentally investigated parameter space provides controlled size adjustment without extrapolation, as shown by an in-silico screening step aimed at 450 nm. These results demonstrate how machine learning may be used as a logical pre-experimental screening method in the formulation development of curcumin nanoparticles. |
|
Keywords |
Curcumin-loaded PLGA nanoparticles, Random Forest regression, Particle size prediction, In-silico formulation screening |
|
URL |
