Article Preview
TopIntroduction
Nanotechnology is a branch of science and engineering that involves the manipulation of matter with at least one dimension sized from 1 to 100 nanometers (Jeevanandam, Barhoum, Chan, Dufresne, & Danquah, 2018). Nanomaterials or nanoparticles, due to their small sizes in nanometer range have fascinating properties that widely vary from the corresponding bulk material. Thus, in recent years, the nanomaterials have gained a prominent status due to their diverse applications in several fields such as food packaging, medicines, electronics, consumer products, optical devices, etc. (Özogul, McClements, Kosker, Durmus, &Ucar, 2019). However, along with the exciting opportunities, there have been growing concerns regarding the risks of nanomaterials on the environment as well as human health (Jeevanandam et al., 2018). For instance, the nanomaterials used in consumer products like pharmaceuticals, cosmetics, powdered food, etc. are anticipated to end up into aquatic, and/or terrestrial environments, where their behavior, toxic effects, and fate are still not completely predictable (Handy & Shaw, 2007; Maynard et al., 2006).
Although the experimental techniques like high throughput screening (HTS) allow to execute large batteries of toxicity assays, they are expensive as well as time-consuming. Further, it is becoming more and more tedious to manage the experimental determination of toxicity for progressively growing nano-particles space considering the possible combinations of nanoparticles showing different sizes, shapes, and chemical compositions, etc. In this scenario, the quantitative structure-activity/toxicity relationship (QSAR/QSTR) models (Puzyn et al., 2011; Roy, Kar, & Das, 2015) play an important role to enable cost-effective means of determining or screening potential nano hazards, and thus helps in reducing the burden of in vitro/in vivo assays. Moreover, QSAR models also help in understanding the structural features or factors that are responsible for their toxicity. Several nano-QSAR models have been already reported ((Gajewicz et al., 2015; Kar, Gajewicz, Puzyn, Roy, & Leszczynski, 2014; Mu et al., 2016; Pathakoti, Huang, Watts, He, & Hwang, 2014; Puzyn et al., 2011; Singh, Gupta, Kumar, & Mohan, 2014; Sizochenko et al., 2014; Toropov et al., 2012)) for predicting the toxicity of metal oxide nanoparticles. Although the prediction quality of these nano-QSTR models were within the acceptable range, these models were directed to single target only. Thus, it should be noted that the conventional QSAR models have the ability to predict the toxicity of nanoparticles against only one biological target and may not always take into consideration several important experimental parameters/conditions such as cell-line used, nanoparticle core size, shape, time of exposure, concentration exposed, etc. Therefore, in recent years, several researchers are focusing on developing multi-tasking QSAR models that are capable of handling multiple biological targets and/or multiple experimental conditions simultaneously. Several new approaches were reported in the literature, for instance, some authors have reported mtk-QSTR-perturbation modeling technique to develop multitasking nano-QSAR models to predict ecotoxicity, genotoxicity and/or cytotoxicity of nanoparticles (Halder, Melo, & Cordeiro, 2020; Kleandrova et al., 2014; Luan et al., 2014). While in another study (Basant & Gupta, 2017) the authors have modified the traditional QSAR methodology and reported an optimal multi target-QSTR model having a functional relationship between four different toxicity endpoints and corresponding predictors. Moreover, Choi et. al., (2018) reported a new methodology to develop a generalized nano-QSAR model by combining physicochemical properties, quantum-mechanical parameters, and different biological experimental conditions as descriptors/attributes, etc.