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.

Multitask Deep Learning Enabling a Synergy for Cadmium and Methane Mitigation with Biochar Amendments in Paddy Soils.

Biochar has demonstrated significant promise in addressing heavy metal contamination and methane (CH4) emissions in paddy soils; however, achieving a synergy between these two goals is challenging due to various variables, including the characteristics of biochar and soil properties that influence biochar's performance. Here, we successfully developed an interpretable multitask deep learning (MTDL) model by employing a tensor tracking paradigm to facilitate parameter sharing between two separate data sets, enabling a synergy between Cd and CH4 mitigation with biochar amendments. The characteristics of biochar contribute similar weightings of 67.9% and 62.5% to Cd and CH4 mitigation, respectively, but their relative importance in determining biochar's performance varies significantly. Notably, this MTDL model excels in custom-tailoring biochar to synergistically mitigate Cd and CH4 in paddy soils across a wide geographic range, surpassing traditional machine learning models. Our findings deepen our understanding of the interactive effects of Cd and CH4 mitigation with biochar amendments in paddy soils, and they also potentially extend the application of artificial intelligence in sustainable environmental remediation, especially when dealing with multiple objectives.

Impacts of biochar aging on its interactions with As(III) and the combined cytotoxicity.

Despite the extensive use of biochar (BC) in soil and aqueous media for pollutant immobilization, the environmental behaviors and health risks of aged BC with multiple pollutants, especially with metal ions possessing various valence states, remain unexplored. Here, we prepared fresh banana peel BC (BP-BC) and aged BP-BCs by acidification (ABP-BC) and oxidation (OBP-BC). ABP-BC was then chosen to explore its environmental behaviors (i.e., adsorption, desorption, and arsenic valence transfer) towards As(III)-Cu(II) and the combined cytotoxicity of BCs with As(III)-Cu(II) was investigated in Human Gastric epithelium cells (GES-1). Our results demonstrate that the aging process notably alters the physicochemical properties of BP-BC, including surface morphology, elemental composition, and surface functional groups, which are key factors affecting the long-term environmental behaviors of BC with As(III)/Cu(II). Specifically, the aging process significantly enhanced the adsorption of As(III) on BC but reduced the adsorption of Cu(II). Although the oxidation of As(III) to As(V) did not change much, the aging process improved the stability of ABP-BC-metal ion complexes, alleviating the release of As(III) in acidic solution. Consequently, the combined cytotoxicity induced by ABP-BC-As(III)-Cu(II) was reduced compared to BP-BC-As(III)-Cu(II). The study highlights the critical roles of the aging process in regulating the As(III) adsorption/desorption dynamics on BCs and their combined cytotoxicity in the presence of multiple metal ions.

Neurotoxicity of the Cu(OH)2 Nanopesticide through Perturbing Multiple Neurotransmitter Pathways in Developing Zebrafish.

The copper hydroxide [Cu(OH)2] nanopesticide is an emerging agricultural chemical that can negatively impact aquatic organisms. This study evaluated the behavioral changes of zebrafish larvae exposed to the Cu(OH)2 nanopesticide and assessed its potential to induce neurotoxicity. Metabolomic and transcriptomic profiling was also conducted to uncover the molecular mechanisms related to potential neurotoxicity. The Cu(OH)2 nanopesticide at 100 µg/L induced zebrafish hypoactivity, dark avoidance, and response to the light stimulus, suggestive of neurotoxic effects. Altered neurotransmitter-related pathways (serotoninergic, dopaminergic, glutamatergic, GABAergic) and reduction of serotonin (5-HT), dopamine (DA), glutamate (GLU), γ-aminobutyric acid (GABA), and several of their precursors and metabolites were noted following metabolomic and transcriptomic analyses. Differentially expressed genes (DEGs) were associated with the synthesis, transport, receptor binding, and metabolism of 5-HT, DA, GLU, and GABA. Transcripts (or protein levels) related to neurotransmitter receptors for 5-HT, DA, GLU, and GABA and enzymes for the synthesis of GLU and GABA were downregulated. Effects on both the glutamatergic and GABAergic pathways in zebrafish were specific to the nanopesticide and differed from those in fish exposed to copper ions. Taken together, the Cu(OH)2 nanopesticide induced developmental neurotoxicity in zebrafish by inhibiting several neurotransmitter-related pathways. This study presented a model for Cu(OH)2 nanopesticide-induced neurotoxicity in developing zebrafish that can inform ecological risk assessments.

Polystyrene nanoplastic exposure induces excessive mitophagy by activating AMPK/ULK1 pathway in differentiated SH-SY5Y cells and dopaminergic neurons in vivo.

BACKGROUND: Microplastics and nanoplastics (MNPs) are emerging environmental contaminants detected in human samples, and have raised concerns regarding their potential risks to human health, particularly neurotoxicity. This study aimed to investigate the deleterious effects of polystyrene nanoplastics (PS-NPs, 50 nm) and understand their mechanisms in inducing Parkinson's disease (PD)-like neurodegeneration, along with exploring preventive strategies. METHODS: Following exposure to PS-NPs (0.5-500 µg/mL), we assessed cytotoxicity, mitochondrial integrity, ATP levels, and mitochondrial respiration in dopaminergic-differentiated SH-SY5Y cells. Molecular docking and dynamic simulations explored PS-NPs' interactions with mitochondrial complexes. We further probed mitophagy's pivotal role in PS-NP-induced mitochondrial damage and examined melatonin's ameliorative potential in vitro. We validated melatonin's intervention (intraperitoneal, 10 mg/kg/d) in C57BL/6 J mice exposed to 250 mg/kg/d of PS-NPs for 28 days. RESULTS: In our in vitro experiments, we observed PS-NP accumulation in cells, including mitochondria, leading to cell toxicity and reduced viability. Notably, antioxidant treatment failed to fully rescue viability, suggesting reactive oxygen species (ROS)-independent cytotoxicity. PS-NPs caused significant mitochondrial damage, characterized by altered morphology, reduced mitochondrial membrane potential, and decreased ATP production. Subsequent investigations pointed to PS-NP-induced disruption of mitochondrial respiration, potentially through interference with complex I (CI), a concept supported by molecular docking studies highlighting the influence of PS-NPs on CI. Rescue experiments using an AMPK pathway inhibitor (compound C) and an autophagy inhibitor (3-methyladenine) revealed that excessive mitophagy was induced through AMPK/ULK1 pathway activation, worsening mitochondrial damage and subsequent cell death in differentiated SH-SY5Y cells. Notably, we identified melatonin as a potential protective agent, capable of alleviating PS-NP-induced mitochondrial dysfunction. Lastly, our in vivo experiments demonstrated that melatonin could mitigate dopaminergic neuron loss and motor impairments by restoring mitophagy regulation in mice. CONCLUSIONS: Our study demonstrated that PS-NPs disrupt mitochondrial function by affecting CI, leading to excessive mitophagy through the AMPK/ULK1 pathway, causing dopaminergic neuron death. Melatonin can counteract PS-NP-induced mitochondrial dysfunction and motor impairments by regulating mitochondrial autophagy. These findings offer novel insights into the MNP-induced PD-like neurodegenerative mechanisms, and highlight melatonin's protective potential in mitigating the MNP's environmental risk.

Computer-Aided Discovery of Potent Broad-Spectrum Vaccine Adjuvants.

Adjuvants stimulate the immune system to vigorously respond to a vaccine. While current adjuvants such as aluminum salts and oil-in-water emulsions have been used for decades, they do not generate broad and long-lasting responses in many vaccines. Consequently, more potent adjuvants are needed. Here, using computer-aided molecule design and machine learning, we discovered 2 new, broad-spectrum adjuvants that can boost vaccine responses. Our library containing 46 toll-like receptor (TLR)-targeting agonist ligands were assembled on Au nanoparticles. Comprehensive in vitro, ex vivo and in vivo studies showed both leads promoted dendritic cell activation via multiple TLRs and enhanced antigen presentation to T cells. When used together with tumor-specific antigens to immunize mice against B16-OVA melanoma and 4T1-PD1 breast cancer, both adjuvants unleashed strong immune responses that suppressed tumor growth and lung metastases. Our results show computer-aided design and screening can rapidly uncover potent adjuvants for tackling waning immunity in current vaccines.

Emerging impacts of ionic liquids on eco-environmental safety and human health.

Owing to their unique physicochemical properties, ionic liquids (ILs) have been rapidly applied in diverse areas, such as organic synthesis, electrochemistry, analytical chemistry, functional materials, pharmaceutics, and biomedicine. The increase in the production and application of ILs has resulted in their release into aquatic and terrestrial environments. Because of their low vapor pressure, ILs cause very little pollution in the atmosphere compared to organic solvents. However, ILs are highly persistent in aquatic and terrestrial environments due to their stability, and therefore, potentially threaten the safety of eco-environments and human health. Specifically, the environmental translocation and retention of ILs, or their accumulation in organisms, are all related to their physiochemical properties, such as hydrophobicity. Based on results of ecotoxicity, cytotoxicity, and toxicity in mammalian models, the mechanisms involved in IL-induced toxicity include damage of cell membranes and induction of oxidative stress. Recently, artificial intelligence and machine learning techniques have been used in mining and modeling toxicity data to make meaningful predictions. Major future challenges are also discussed. This review will accelerate our understanding of the safety issues of ILs and serve as a guideline for the design of the next generation of ILs.

Comprehensive Interrogation on Acetylcholinesterase Inhibition by Ionic Liquids Using Machine Learning and Molecular Modeling.

Quantitative structure-activity relationship (QSAR) modeling can be used to predict the toxicity of ionic liquids (ILs), but most QSAR models have been constructed by arbitrarily selecting one machine learning method and ignored the overall interactions between ILs and biological systems, such as proteins. In order to obtain more reliable and interpretable QSAR models and reveal the related molecular mechanism, we performed a systematic analysis of acetylcholinesterase (AChE) inhibition by 153 ILs using machine learning and molecular modeling. Our results showed that more reliable and stable QSAR models (R2 > 0.85 for both cross-validation and external validation) were obtained by combining the results from multiple machine learning approaches. In addition, molecular docking results revealed that the cations and organic anions of ILs bound to specific amino acid residues of AChE through noncovalent interactions such as π interactions and hydrogen bonds. The calculation results of binding free energy showed that an electrostatic interaction (ΔEele < -285 kJ/mol) was the main driving force for the binding of ILs to AChE. The overall findings from this investigation demonstrate that a systematic approach is much more convincing. Future research in this direction will help design the next generation of biosafe ILs.

Elucidation of the Critical Role of Core Materials in PM2.5-Induced Cytotoxicity by Interrogating Silica- and Carbon-Based Model PM2.5 Particle Libraries.

An insoluble core with adsorbed pollutants constitutes the most toxic part of PM2.5 particles. However, the toxicological difference between carbon and silica cores remains unknown. Here, we employed 32-membered carbon- and silica-based model PM2.5 libraries that each was loaded with four toxic airborne pollutants including Cr(VI), As(III), Pb2+, and BaP in all possible combinations to explore their contributions to cytotoxicity in normal human bronchial cells. The following three crucial findings were revealed: (1) more adsorption of polar pollutants in a silica core (such as Cr(VI), As(III), and Pb2+) and nonpolar ones in a carbon core (such as BaP); (2) about 41% more cell uptake of carbon- than silica-based particles; and (3) about 59% less toxicity in silica- than carbon-based particles when pollutants other than Cr(VI) were loaded. This was reversed after Cr(VI) loading (silica particles were 56% more toxic). The difference maker is that compared to stable silica, carbon particles reduce Cr(VI) to less toxic Cr(III). Our findings highlight the different roles of carbon and silica cores in inducing health risks of PM2.5 particles.

Cytotoxicity Induction by the Oxidative Reactivity of Nanoparticles Revealed by a Combinatorial GNP Library with Diverse Redox Properties.

It is crucial to establish relationship between nanoparticle structures (or properties) and nanotoxicity. Previous investigations have shown that a nanoparticle's size, shape, surface and core materials all impact its toxicity. However, the relationship between the redox property of nanoparticles and their toxicity has not been established when all other nanoparticle properties are identical. Here, by synthesizing an 80-membered combinatorial gold nanoparticle (GNP) library with diverse redox properties, we systematically explored this causal relationship. The compelling results revealed that the oxidative reactivity of GNPs, rather than their other physicochemical properties, directly caused cytotoxicity via induction of cellular oxidative stress. Our results show that the redox diversity of nanoparticles is regulated by GNPs modified with redox reactive ligands.

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.

Regulation of Aryl Hydrocarbon Receptor Signaling Pathway and Dioxin Toxicity by Novel Agonists and Antagonists.

Dioxins, mostly through activation of aryl hydrocarbon receptor (AhR), are potent toxic substances widely distributed in the environment, while moderated suppression of AhR also exhibits anti-tumor effects. Therefore, the proper modulation of AhR activity may counteract AhR-mediated toxicities and certain diseases. In this investigation, we identified several novel AhR moderate agonists and antagonists using chemical biology approaches. The mechanisms and mode of interactions with AhR by these hits were also revealed using both experimental and computational studies. The newly identified AhR moderate agonists and antagonists were predicted to bind to AhR and modulate AhR signaling. The structure-activity relationships of moderate agonists and antagonists and their unique binding features with AhR have created a solid framework for further optimization of the next generation of AhR modulators.

Analysis of model PM2.5-induced inflammation and cytotoxicity by the combination of a virtual carbon nanoparticle library and computational modeling.

Health risks induced by PM2.5 have become one of the major concerns among living populations, especially in regions facing serious pollution such as China and India. Furthermore, the composition of PM2.5 is complex and it also varies with time and locations. To facilitate our understanding of PM2.5-induced toxicity, a predictive modeling framework was developed in the present study. The core of this study was 1) to construct a virtual carbon nanoparticle library based on the experimental data to simulate the PM2.5 structures; 2) to quantify the nanoparticle structures by novel nanodescriptors; and 3) to perform computational modeling for critical toxicity endpoints. The virtual carbon nanoparticle library was developed to represent the nanostructures of 20 carbon nanoparticles, which were synthesized to simulate PM2.5 structures and tested for potential health risks. Based on the calculated nanodescriptors from virtual carbon nanoparticles, quantitative nanostructure-activity relationship (QNAR) models were developed to predict cytotoxicity and four different inflammatory responses induced by model PM2.5. The high predictability (R2 > 0.65 for leave-one-out validations) of the resulted consensus models indicated that this approach could be a universal tool to predict and analyze the potential toxicity of model PM2.5, ultimately understanding and evaluating the ambient PM2.5-induced toxicity.

Transport behavior of a single Ca(2+), K(+), and Na(+) in a water-filled transmembrane cyclic peptide nanotube.

Molecular dynamics simulations have been performed to investigate the transport properties of a single Ca(2+), K(+), and Na(+) in a water-filled transmembrane cyclic peptide nanotube (CPNT). Two transmembrane CPNTs, i.e., 8×(WL)n=4,5/POPE (with uniform lengths but various radii), were applied to clarify the dependence of ionic transport properties on the channel radius. A huge energy barrier keeps Ca(2+) out of the octa-CPNT, while Na(+) and K(+) can be trapped in two CPNTs. The dominant electrostatic interaction of a cation with water molecules leads to a high distribution of channel water around the cation and D-defects in the first and last gaps, and significantly reduces the axial diffusion of channel water. Water-bridged interactions were mostly found between the artificially introduced Ca(2+) and the framework of the octa-CPNT, and direct coordinations with the tube wall mostly occur for K(+) in the octa-CPNT. A cation may drift rapidly or behave lazily in a CPNT. K(+) behaves most actively and can visit the whole deca-CPNT quickly. The first solvation shells of Ca(2+) and Na(+) are basically saturated in two CPNTs, while the hydration of K(+) is incomplete in the octa-CPNT. The solvation structure of Ca(2+) in the octa-CPNT is most stable, while that of K(+) in the deca-CPNT is most labile. Increasing the channel radius induces numerous interchange attempts between the first-shell water molecules of a cation and the ones in the outer region, especially for the K(+) system.

Tilt behavior of an octa-peptide nanotube in POPE and affects on the transport characteristics of channel water.

MD simulations have been carried out for studying the tilt behaviors of the 8 × (WL)4-CPNT embedded in the POPE bilayer in a water environment and under an anhydrous condition, respectively. Besides, the dependences of the transport characteristics of channel water on the extent of the CPNT tilt were explored. The results indicate that the presence of water may exacerbate the CPNT tilt but plays an important role in maintaining the integrity of the CPNT by forming water bridges in the end-gaps. Cation-π interactions between the indole rings of Trp residues and lipid headgroups are the major factor causing the CPNT tilt under an anhydrous condition, while H-bonded interactions between water molecules and the indole rings are primary in a water environment. The dipole orientations of channel water molecules except those in the last end-gap are gradually oriented and eventually only take "+dipole" states under 20° of the CPNT tilt. The average residence time of channel water gradually increases with the intensifying of the CPNT tilt.

Dynamic behaviors and transport properties of ethanol molecules in transmembrane cyclic peptide nanotubes.

Classical molecular dynamics simulations have been performed to investigate the dynamic behaviors and transport properties of ethanol molecules in transmembrane cyclic peptide nanotubes (CPNTs) with various radii, i.e., 8×(WL¯)n=3,4,5/POPE. The results show that ethanol molecules spontaneously fill the octa- and deca-CPNTs, but not the hexa-CPNT. In the octa-CPNT, ethanol molecules are trapped at individual gaps with their carbon skeletons perpendicular to the tube axis and hydroxyl groups towards the tube wall, forming a broken single-file chain. As the channel radius increases, ethanol molecules inside the deca-CPNT tend to form a tubular layer and the hydroxyl groups mainly stretch towards the tube axis. Computations of diffusion coefficients indicate that ethanol molecules in the octa-CPNT nearly lost their diffusion abilities, while those in the deca-CPNT diffuse as 4.5 times as in a (8, 8) carbon nanotube with a similar tube diameter. The osmotic and diffusion permeabilities (pf and pd, respectively) of the octa- and deca-CPNTs transporting ethanol were deduced for the first time. The distributions of the gauche and trans conformers of ethanol molecules in two CPNTs are quite similar, both with approximately 57% gauche conformers. The non-bonded interactions of channel ethanol with a CPNT wall and surrounding ethanol were explored. The potential of mean force elucidates the mechanism underlying the transporting characteristics of channel ethanol in a transmembrane CPNT.

Effect-directed analysis of androgenic compounds from sewage sludges in China.

The safety of municipal sewage sludge has raised great concerns because of the accumulation of large-scale endocrine disrupting chemicals in the sludge during wastewater treatment. The presence of contaminants in sludge can cause secondary pollution owing to inappropriate disposal mechanisms, posing potential risks to the environment and human health. Effect-directed analysis (EDA), involving an androgen receptor (AR) reporter gene bioassay, fractionation, and suspect and nontarget chemical analysis, were applied to identify causal AR agonists in sludge; 20 of the 30 sludge extracts exhibited significant androgenic activity. Among these, the extracts from Yinchuan, Kunming, and Shijiazhuang, which held the most polluted AR agonistic activities were prepared for extensive EDA, with the dihydrotestosterone (DHT)-equivalency of 2.5 - 4.5 ng DHT/g of sludge. Seven androgens, namely boldione, androstenedione, testosterone, megestrol, progesterone, and testosterone isocaproate, were identified in these strongest sludges together, along with testosterone cypionate, first reported in sludge media. These identified androgens together accounted for 55 %, 87 %, and 52 % of the effects on the sludge from Yinchuan, Shijiazhuang, and Kunming, respectively. This study elucidates the causative androgenic compounds in sewage sludge and provides a valuable reference for monitoring and managing androgens in wastewater treatment.

Unraveling the joint toxicity of transition-metal dichalcogenides and per- and polyfluoroalkyl substances in aqueous mediums by experimentation, machine learning and molecular dynamics.

The environmental fate of transition-metal dichalcogenides (TMDCs) may be further complicated by interacting with existing pollutants, especially per- and polyfluoroalkyl substances (PFAS). However, due to their sheer volume, it is impossible to explore all possible interactions by simply utilizing experimental methods. Herein, we used two model TMDC nanosheets, molybdenum disulfide (MoS2) and tungsten disulfide (WS2), and seven PFAS to explore their interactions and subsequent impacts on model cell lines and zebrafish. Utilizing experimental methods and machine learning approaches, we showed that TMDCs-PFAS interactions can pose unique challenges due to their interaction-specific toxicity niches towards cell lines. Further in vivo experiments, together with molecular dynamics simulation, suggested that TMDCs-PFAS interactions in aqueous environments significantly increased their bioaccumulation in zebrafish towards different target organs, mostly due to the differences in loading PFAS. Such enhanced bioaccumulation increased the oxidative stress in zebrafish liver and intestine, as demonstrated by the increased reactive oxygen species (ROS) level and other enzyme activities, which eventually led to obvious histopathological alterations in the liver and intestine. Our study highlights the importance of exploring interactions between emerging and existing contaminants with state-of-art techniques in aqueous environments and its significance in safeguarding aquatic environment health.

Surface functional groups and biofilm formation on microplastics: Environmental implications.

Microplastics (MPs) contamination is becoming a significant environmental issue, as the widespread omnipresence of MPs can cause many adverse consequences for both ecological systems and humans. Contrary to what is commonly thought, the toxicity-inducing MPs are not the original pristine plastics; rather, they are completely transformed through various surface functional groups and aggressive biofilm formation on MPs via aging or weathering processes. Therefore, understanding the impacts of MPs' surface functional groups and biofilm formation on biogeochemical processes, such as environmental fate, transport, and toxicity, is crucial. In this review, we present a comprehensive summary of the distinctive impact that surface functional groups and biofilm formation of MPs have on their significant biogeochemical behavior in various environmental media, as well as their toxicity and biological effects. We place emphasis on the role of surface functional groups and biofilm formation as a means of influencing the biogeochemical processes of MPs. This includes their effects on pollutant fate and element cycling, which in turn impacts the aggregation, transport, and toxicity of MPs. Ultimately, future research studies and tactics are needed to improve our understanding of the biogeochemical processes that are influenced by the surface functional groups and biofilm formation of MPs.

Implementing comprehensive machine learning models of multispecies toxicity assessment to improve regulation of organic compounds.

Machine learning has made significant progress in assessing the risk associated with hazardous chemicals. However, most models were constructed by randomly selecting one algorithm and one toxicity endpoint towards single species, which may cause biased regulation of chemicals. In the present study, we implemented comprehensive prediction models involving multiple advanced machine learning and end-to-end deep learning to assess the aquatic toxicity of chemicals. The generated optimal models accurately unravel the quantitative structure-toxicity relationships, with the correlation coefficients of all training sets from 0.59 to 0.81 and of the test sets from 0.56 to 0.83. For each chemical, its ecological risk was determined from the toxicity information towards multiple species. The results also revealed the toxicity mechanism of chemicals was species sensitivity, and the high-level organisms were faced with more serious side effects from hazardous substances. The proposed approach was finally applied to screen over 16,000 compounds and identify high-risk chemicals. We believe that the current approach can provide a useful tool for predicting the toxicity of diverse organic chemicals and help regulatory authorities make more reasonable decisions.