Commercial use of nanoparticles is on the rise, but our understanding of how these particles behave in living organisms is still in its infancy. Their worth, and their potential danger, stem from the fact that the characteristics of the nanophase are very different from the bulk properties of the same material. Nanoparticle uptake, distribution, modification, and potential toxicity must all be investigated experimentally. Predictive models would be very helpful, especially for helping regulators reduce health and environmental hazards, but they are expensive and time-consuming to develop. Three datasets were modelled, one of which contained nanoparticles, with the use of sparse machine learning algorithms and Bayesian neural networks. In the initial phase of experimentation, researchers employed: Pancreatic cancer cells, Human Umbilical Vein Endothelial Cells (HUVEC) and three macrophage or macrophage-like cell lines. These cells were exposed to iron oxide nanoparticles coated with a diverse set of 108 compounds, coatings, fifty two nanoparticles with different core materials, and surface changes were added. These nanoparticles underwent comprehensive analysis using four parameters, including size, relaxivity, and zeta potential. Additionally, their impact on cell lines was evaluated through a total of eight measurements, comprising four biological assessments per cell line, each conducted at four different dosage levels. In the third batch of data, 80 different small compounds were used to modify gold nanoparticles and study their biological effects. The biological outcomes that were modelled were binding to AChE and nonspecific binding. Nanoparticles’ biological impacts were modelled with help from chemical descriptors for the substances that covered their surfaces. Most biological outcomes were modelled with high statistical quality. These proof-of-concept models demonstrate that it is feasible, with the help of current modelling techniques, to simulate the impacts of nanomaterials on living organisms.

PDF (Download Full text)

Download Citation (XML)