A further development of the QNAR model to predict the cellular uptake of nanoparticles by pancreatic cancer cells

Identification of structural features of surface modifiers in engineered nanostructured metal oxides regarding cell uptake through ML-based classification

Nanoparticles (NPs) are considered as versatile tools in various fields including medicine, electronics, and environmental science. Understanding the structural aspects of surface modifiers in nanoparticles that govern their cellular uptake is crucial for optimizing their efficacy and minimizing potential cytotoxicity. The cellular uptake is influenced by multiple factors, namely, size, shape, and surface charge of NPs, as well as their surface functionalization. In the current study, classification-based ML models (i.e., Bayesian classification, random forest, support vector classifier, and linear discriminant analysis) have been developed to identify the features/fingerprints that significantly contribute to the cellular uptake of ENMOs in multiple cell types, including pancreatic cancer cells (PaCa2), human endothelial cells (HUVEC), and human macrophage cells (U937). The best models have been identified for each cell type and analyzed to detect the structural fingerprints/features governing the cellular uptake of ENMOs. The study will direct scientists in the design of ENMOs of higher cellular uptake efficiency for better therapeutic response.

Unveiling protein corona composition: predicting with resampling embedding and machine learning

Biomaterials with surface nanostructures effectively enhance protein secretion and stimulate tissue regeneration. When nanoparticles (NPs) enter the living system, they quickly interact with proteins in the body fluid, forming the protein corona (PC). The accurate prediction of the PC composition is critical for analyzing the osteoinductivity of biomaterials and guiding the reverse design of NPs. However, achieving accurate predictions remains a significant challenge. Although several machine learning (ML) models like Random Forest (RF) have been used for PC prediction, they often fail to consider the extreme values in the abundance region of PC absorption and struggle to improve accuracy due to the imbalanced data distribution. In this study, resampling embedding was introduced to resolve the issue of imbalanced distribution in PC data. Various ML models were evaluated, and RF model was finally used for prediction, and good correlation coefficient (R2) and root-mean-square deviation (RMSE) values were obtained. Our ablation experiments demonstrated that the proposed method achieved an R2 of 0.68, indicating an improvement of approximately 10%, and an RMSE of 0.90, representing a reduction of approximately 10%. Furthermore, through the verification of label-free quantification of four NPs: hydroxyapatite (HA), titanium dioxide (TiO2), silicon dioxide (SiO2) and silver (Ag), and we achieved a prediction performance with an R2 value >0.70 using Random Oversampling. Additionally, the feature analysis revealed that the composition of the PC is most significantly influenced by the incubation plasma concentration, PDI and surface modification.

Converting Nanotoxicity Data to Information Using Artificial Intelligence and Simulation

Decades of nanotoxicology research have generated extensive and diverse data sets. However, data is not equal to information. The question is how to extract critical information buried in vast data streams. Here we show that artificial intelligence (AI) and molecular simulation play key roles in transforming nanotoxicity data into critical information, i.e., constructing the quantitative nanostructure (physicochemical properties)-toxicity relationships, and elucidating the toxicity-related molecular mechanisms. For AI and molecular simulation to realize their full impacts in this mission, several obstacles must be overcome. These include the paucity of high-quality nanomaterials (NMs) and standardized nanotoxicity data, the lack of model-friendly databases, the scarcity of specific and universal nanodescriptors, and the inability to simulate NMs at realistic spatial and temporal scales. This review provides a comprehensive and representative, but not exhaustive, summary of the current capability gaps and tools required to fill these formidable gaps. Specifically, we discuss the applications of AI and molecular simulation, which can address the large-scale data challenge for nanotoxicology research. The need for model-friendly nanotoxicity databases, powerful nanodescriptors, new modeling approaches, molecular mechanism analysis, and design of the next-generation NMs are also critically discussed. Finally, we provide a perspective on future trends and challenges.

Application of Computing as a High-Practicability and -Efficiency Auxiliary Tool in Nanodrugs Discovery

Research and development (R&D) of nanodrugs is a long, complex and uncertain process. Since the 1960s, computing has been used as an auxiliary tool in the field of drug discovery. Many cases have proven the practicability and efficiency of computing in drug discovery. Over the past decade, computing, especially model prediction and molecular simulation, has been gradually applied to nanodrug R&D, providing substantive solutions to many problems. Computing has made important contributions to promoting data-driven decision-making and reducing failure rates and time costs in discovery and development of nanodrugs. However, there are still a few articles to examine, and it is necessary to summarize the development of the research direction. In the review, we summarize application of computing in various stages of nanodrug R&D, including physicochemical properties and biological activities prediction, pharmacokinetics analysis, toxicological assessment and other related applications. Moreover, current challenges and future perspectives of the computing methods are also discussed, with a view to help computing become a high-practicability and -efficiency auxiliary tool in nanodrugs discovery and development.

Machine Learning for Evaluating the Cytotoxicity of Mixtures of Nano-TiO2 and Heavy Metals: QSAR Model Apply Random Forest Algorithm after Clustering Analysis

With the development and application of nanomaterials, their impact on the environment and organisms has attracted attention. As a common nanomaterial, nano-titanium dioxide (nano-TiO₂) has adsorption properties to heavy metals in the environment. Quantitative structure-activity relationship (QSAR) is often used to predict the cytotoxicity of a single substance. However, there is little research on the toxicity of interaction between nanomaterials and other substances. In this study, we exposed human renal cortex proximal tubule epithelial (HK-2) cells to mixtures of eight heavy metals with nano-TiO₂, measured absorbance values by CCK-8, and calculated cell viability. PLS and two ensemble learning algorithms are used to build multiple QSAR models for data sets, and the test set R² is increased from 0.38 to 0.78 and 0.85, and RMSE is decreased from 0.18 to 0.12 and 0.10. After selecting the better random forest algorithm, the K-means clustering algorithm is used to continue to optimize the model, increasing the test set R² to 0.95 and decreasing the RMSE to 0.08 and 0.06. As a reliable machine algorithm, random forest can be used to predict the toxicity of the mixture of nano-metal oxides and heavy metals. The cluster analysis can effectively improve the stability and predictability of the model, and provide a new idea for the prediction of cytotoxicity model in the future.

QNAR modeling of cytotoxicity of mixing nano-TiO2 and heavy metals

The Quantitative Structure-Activity Relationship (QSAR) has been used to investigate organic mixtures but QSAR in the nanomaterial field (QNAR) is still new. Toxicity is a result of the interaction of many substances. QNAR research focuses on a single nanomaterial in the long-term. It is difficult to find an appropriate descriptor to build a model due to the complexity of the mixture. Here, we attempt to build a QNAR model to predict cell viability for HK-2 cells exposed to a mixture containing nano-TiO₂ and heavy metals. HK-2 cells were exposed to four groups of mixtures containing heavy-metals and nanomaterials and CCK8 was added to obtain the number of living cells. At the same time, ROS was investigated to study this mechanism. Each descriptor of the components and mixtures were obtained using the formula D_mix= [Formula: see text] respectively. We used the Multiple Partial Least Squares Regression (PLS) and Random Forest Regression (RF) to build a QNAR model. Both models reliably predict and assess viability of HK-2 cells exposed to the mixture. The RF model showed greater stability and higher precision in toxicity predictability and can be applied to environmental nano-toxicology.

Interaction of nanoparticles with biological macromolecules: a review of molecular docking studies

The high frequency of using engineered nanoparticles in various medical applications entails a deep understanding of their interaction with biological macromolecules. Molecular docking simulation is now widely used to study the binding of different types of nanoparticles with proteins and nucleic acids. This helps not only in understanding the mechanism of their biological action but also in predicting any potential toxicity. In this review, the computational techniques used in studying the nanoparticles interaction with biological macromolecules are covered. Then, a comprehensive overview of the docking studies performed on various types of nanoparticles will be offered. The implication of these predicted interactions in the biological activity and/or toxicity is also discussed for each type of nanoparticles.

Virtual Molecular Projections and Convolutional Neural Networks for the End-to-End Modeling of Nanoparticle Activities and Properties

Digitalizing complex nanostructures into data structures suitable for machine learning modeling without losing nanostructure information has been a major challenge. Deep learning frameworks, particularly convolutional neural networks (CNNs), are especially adept at handling multidimensional and complex inputs. In this study, CNNs were applied for the modeling of nanoparticle activities exclusively from nanostructures. The nanostructures were represented by virtual molecular projections, a multidimensional digitalization of nanostructures, and used as input data to train CNNs. To this end, 77 nanoparticles with various activities and/or physicochemical property results were used for modeling. The resulting CNN model predictions show high correlations with the experimental results. An analysis of a trained CNN quantitatively showed that neurons were able to recognize distinct nanostructure features critical to activities and physicochemical properties. This "end-to-end" deep learning approach is well suited to digitalize complex nanostructures for data-driven machine learning modeling and can be broadly applied to rationally design nanoparticles with desired activities.

New Relevant Descriptor of Linear QNAR Models for Toxicity Assessment of Silver Nanoparticles

The use of silver nanoparticles (NPs) in medical, industrial and agricultural fields is becoming more widespread every year. This leads to an increasing number of experimental toxicological and microbiological studies of silver NPs aimed at establishing the risk-benefit ratio for their application. The following key parameters affecting the biological activity of silver dispersions are traditionally taken into consideration: mean diameter of NPs, surface potential of NPs and equilibrium concentration of Ag+. These characteristics are mainly predetermined by the chemical nature of the capping agent used for stabilization. However, the extent to which they influence the biological activity and the toxicity of silver NPs varies greatly. In this work, dispersions of silver NPs stabilized with a wide array of substances of different chemical nature were used for quantitative evaluation of whether the various measurable properties of silver NPs fit as descriptors of linear QNAR (quantitative nanostructure-activity relationship) models for silver NP toxicity evaluation with respect to a model eukaryotic microorganism-Saccharomyces cerevisiae yeast cells. It was shown that among the factors that determine silver NP toxicity, the charge of particles, their colloidal stability and the ability to generate Ag+ ions carry more importance than the descriptors related to the particle size. A significant synergistic effect between the ζ-potential and the colloidal stability of silver NPs on their toxicity was also discovered. Following this, a new descriptor has been proposed for the integral characterization of the silver dispersion colloidal stability. According to the obtained data, it can be considered applicable for building QNAR models of higher efficacy. The validity testing of the proposed model for theoretical prediction of silver NP toxicity using a wide range of living organisms has shown that this new descriptor correlates with toxicity much better compared to most traditionally used descriptors. Consequently, it seems promising in terms of being used not only in situations involving the rather narrow array of the objects tested, but also for the construction of silver NP toxicity models with respect to other living organisms.

Importance of Choosing Relevant Biological End Points To Predict Nanoparticle Toxicity with Computational Approaches for Human Health Risk Assessment

Because it is impossible to assess in vitro or in vivo the toxicity of all nanoparticles available on the market on a case-by-case basis, computational approaches have been proposed as useful alternatives to predict in silico the hazard potential of engineered nanoparticles. Despite promising results, a major issue associated with these mathematical models lies in the a priori choice of the physicochemical descriptors and the biological end points. We performed a thorough bibliographic survey on the biological end points used for nanotoxicology purposes and compared them between experimental and computational approaches. They were found to be disparate: while conventional in vitro nanotoxicology assays usually investigate a large array of biological effects using eukaryotic cells (cytotoxicity, pro-inflammatory response, oxidative stress, genotoxicity), computational studies mostly focus on cell viability and also include studies on prokaryotic cells. We may thus wonder the relevance of building complex mathematical models able to predict accurately a biological end point if this latter is not the most relevant to support human health risk assessment. The choice of biological end points clearly deserves to be more carefully discussed. This could bridge the gap between experimental and computational nanotoxicology studies and allow in silico predictive models to reach their full potential.

Universal nanohydrophobicity predictions using virtual nanoparticle library

To facilitate the development of new nanomaterials, especially nanomedicines, a novel computational approach was developed to precisely predict the hydrophobicity of gold nanoparticles (GNPs). The core of this study was to develop a large virtual gold nanoparticle (vGNP) library with computational nanostructure simulations. Based on the vGNP library, a nanohydrophobicity model was developed and then validated against externally synthesized and tested GNPs. This approach and resulted model is an efficient and effective universal tool to visualize and predict critical physicochemical properties of new nanomaterials before synthesis, guiding nanomaterial design.

Estimation of the Toxicity of Different Substituted Aromatic Compounds to the Aquatic Ciliate Tetrahymena pyriformis by QSAR Approach

Nowadays, quantitative structure–activity relationship (QSAR) methods have been widely performed to predict the toxicity of compounds to organisms due to their simplicity, ease of implementation, and low hazards. In this study, to estimate the toxicities of substituted aromatic compounds to Tetrahymena pyriformis, the QSAR models were established by the multiple linear regression (MLR) and radial basis function neural network (RBFNN). Unlike other QSAR studies, according to the difference of functional groups (−NO₂, −X), the whole dataset was divided into three groups and further modeled separately. The statistical characteristics for the models are obtained as the following: MLR: n = 36, R² = 0.829, RMS (root mean square) = 0.192, RBFNN: n = 36, R² = 0.843, RMS = 0.167 for Group 1; MLR: n = 60, R² = 0.803, RMS = 0.222, RBFNN: n = 60, R² = 0.821, RMS = 0.193 for Group 2; MLR: n = 31 R² = 0.852, RMS = 0.192; RBFNN: n = 31, R² = 0.885, RMS = 0.163 for Group 3, respectively. The results were within the acceptable range, and the models were found to be statistically robust with high external predictivity. Moreover, the models also gave some insight on those characteristics of the structures that most affect the toxicity.

Predicting Nano-Bio Interactions by Integrating Nanoparticle Libraries and Quantitative Nanostructure Activity Relationship Modeling

The discovery of biocompatible or bioactive nanoparticles for medicinal applications is an expensive and time-consuming process that may be significantly facilitated by incorporating more rational approaches combining both experimental and computational methods. However, it is currently hindered by two limitations: (1) the lack of high-quality comprehensive data for computational modeling and (2) the lack of an effective modeling method for the complex nanomaterial structures. In this study, we tackled both issues by first synthesizing a large library of nanoparticles and obtained comprehensive data on their characterizations and bioactivities. Meanwhile, we virtually simulated each individual nanoparticle in this library by calculating their nanostructural characteristics and built models that correlate their nanostructure diversity to the corresponding biological activities. The resulting models were then used to predict and design nanoparticles with desired bioactivities. The experimental testing results of the designed nanoparticles were consistent with the model predictions. These findings demonstrate that rational design approaches combining high-quality nanoparticle libraries, big experimental data sets, and intelligent computational models can significantly reduce the efforts and costs of nanomaterial discovery.