Quantitative Prediction of Inorganic Nanomaterial Cellular Toxicity via Machine Learning

Abstract

Organic chemistry has seen colossal progress due to machine learning (ML). However, the translation of artificial intelligence (AI) into materials science is challenging, where biological behavior prediction becomes even more complicated. Nanotoxicity is a critical parameter that describes their interaction with the living organisms screened in every bio‐related research. To prevent excessive experiments, such properties have to be pre‐evaluated. Several existing ML models partially fulfill the gap by predicting whether a nanomaterial is toxic or not. Yet, this binary categorization neglects the concentration dependencies crucial for experimental scientists. Here, an ML‐based approach is proposed to the quantitative prediction of inorganic nanomaterial cytotoxicity achieving the precision expressed by 10‐fold cross‐validation (CV) Q2 = 0.86 with the root mean squared error (RMSE) of 12.2% obtained by the correlation‐based feature selection and grid search‐based model hyperparameters optimization. To provide further model flexibility, quantitative atom property‐based nanomaterial descriptors are introduced allowing the model to extrapolate on unseen samples. Feature importance is calculated to find an interpretable model with optimal decision‐making. These findings allow experimental scientists to perform primary in silico candidate screening and minimize the number of excessive, labor‐intensive experiments enabling the rapid development of nanomaterials for medicinal purposes.

Machine Learning Reinforced Genetic Algorithm for Massive Targeted Discovery of Selectively Cytotoxic Inorganic Nanoparticles

“…The resulting dataset is an expanded version of the dataset that had been successfully used in our previous work. [29] To ensure top performance of ML models, we implemented standard data preprocessing steps including handling of missing values, duplicates, outliers, as well as identification of redundant features and feature selection. For outlier removal, we analyzed distributions of viability, concentration, hydrodynamic diameter, and zeta potential and applied a percentile-based approach (Figure 2c).…”

“…[ 28 ] In our previous work, regression models were developed for quantitative cytotoxicity prediction of inorganic NPs. [ 29 ] Despite high performance metrics frequently reported for predictive models in the past, most models were trained on rather limited data, which inevitably leads to prediction biases and overfitting. Table S1 provides a succinct summary of the data and the corresponding performance metrics in this study compared to the previous works.…”

The nanosafety assessment of ENMs under a dermal exposure scenario: from key molecular events to in silico modeling tools

“…7), nano-(Q)SAR enables the given toxic effects of ENMs to be determined by their 197 With the EU ban on animal testing, (Q)SAR has been extensively used as an alternative approach in mechanistic interpretation, tiered testing, grouping, and ranking the toxic potency of ENMs for risk assessment. Table 5 summarizes recent (Q)SAR studies regarding the cytotoxicity (cellular uptake and HaCaT cell viability) and/or genotoxicity (results of the bacterial reverse mutation test) of MeOx NPs, 112,[198][199][200][201][202][203][204][205][206][207][208][209] carbon nanotubes 210,211 and fullerenes. 212,213 Datasets for (Q)SAR modeling can be obtained from literature, databases or experiments and should contain sufficient chemically diverse data.…”

“…197 With the EU ban on animal testing, (Q)SAR has been extensively used as an alternative approach in mechanistic interpretation, tiered testing, grouping, and ranking the toxic potency of ENMs for risk assessment. Table 5 summarizes recent (Q)SAR studies regarding the cytotoxicity (cellular uptake and HaCaT cell viability) and/or genotoxicity (results of the bacterial reverse mutation test) of MeOx NPs, 112,198–209 carbon nanotubes 210,211 and fullerenes. 212,213…”

Layered nanomaterials for renewable energy generation and storage

“…7b). [200][201][202][203] In this section, we will discuss the current state of applications in optimizing various 2D materials and discuss the problems and opportunities in this rapidly developing field.…”