Academic Journal
Predicting the toxicity of multicomponent complex mixtures using a multi-feature fusion-based machine learning model.
| Τίτλος: | Predicting the toxicity of multicomponent complex mixtures using a multi-feature fusion-based machine learning model. |
|---|---|
| Συγγραφείς: | Zhao Y; Key Laboratory of Water Pollution Control and Wastewater Resource of Anhui Province, College of Environment and Energy Engineering, Anhui Jianzhu University, Hefei, China; Anhui Gaodi Technology Co., LTD, Luan, China., Jia R; Anhui Gaodi Technology Co., LTD, Luan, China., Zhang J; Key Laboratory of Water Pollution Control and Wastewater Resource of Anhui Province, College of Environment and Energy Engineering, Anhui Jianzhu University, Hefei, China; Anhui Gaodi Technology Co., LTD, Luan, China., Xie M; Key Laboratory of Water Pollution Control and Wastewater Resource of Anhui Province, College of Environment and Energy Engineering, Anhui Jianzhu University, Hefei, China., Xie F; Key Laboratory of Water Pollution Control and Wastewater Resource of Anhui Province, College of Environment and Energy Engineering, Anhui Jianzhu University, Hefei, China., Mao W; Key Laboratory of Water Pollution Control and Wastewater Resource of Anhui Province, College of Environment and Energy Engineering, Anhui Jianzhu University, Hefei, China., Zhang J; Key Laboratory of Water Pollution Control and Wastewater Resource of Anhui Province, College of Environment and Energy Engineering, Anhui Jianzhu University, Hefei, China. Electronic address: ginnzy@ahjzu.edu.cn. |
| Πηγή: | Aquatic toxicology (Amsterdam, Netherlands) [Aquat Toxicol] 2026 Oct; Vol. 299, pp. 107967. Date of Electronic Publication: 2026 Aug 03. |
| Τύπος έκδοσης: | Journal Article |
| Γλώσσα: | English |
| Στοιχεία περιοδικού: | Publisher: Elsevier/North Holland Biomedical Press Country of Publication: Netherlands NLM ID: 8500246 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1879-1514 (Electronic) Linking ISSN: 0166445X NLM ISO Abbreviation: Aquat Toxicol Subsets: MEDLINE |
| Imprint Name(s): | Original Publication: Amsterdam, Netherlands : Elsevier/North Holland Biomedical Press, c1981- |
| Ιατρικοί όροι (MeSH): | Chlorella*/drug effects , Water Pollutants, Chemical*/toxicity , Complex Mixtures*/toxicity , Machine Learning*, Predictive Learning Models ; Prediction Algorithms |
| Περίληψη: | Complex chemical mixtures are ubiquitous in aquatic environments posing substantial risks to aquatic organisms and human health, and the prediction of their combined toxicity is the key measure to control their ecological risks. This, however, still remains a major challenge for the two widely used classic standard additive models, concentration addition (CA) and independent action (IA), due to complex interactions, antagonism or synergism, within mixtures. Therefore, to conquer the dilemma, a novel interpretable machine learning model was constructed by integrating the predicted values by CA and IA with molecular descriptors (CIMM) against a freshwater organism Chlorella pyrenoidosa to predict the combined toxicity of complex mixture pollutants. The CIMM model stability was assessed by repeated modeling with ten random seeds, and its interpretability was examined by using mutual information, partial dependence plots and SHAP values. A dual-metric applicability domain was established based on kernel density estimation and k-nearest neighbor distance, and generalization performance was evaluated on a completely independent external validation set. The results showed that the CIMM framework achieved a mean test-set R2 of 0.9080 ± 0.0127 across ten random-seed runs, which was implemented using the seed-specific optimal Random Forest or XGBoost algorithm. Compared to the results predicted by CA alone, CIMM increased R2 by 32.2% and the RMSE decreased by 46.4%, while to those predicted by model using only molecular descriptors, the R2 improved by approximately 8.87% and the RMSE decreased by 25.38%. CA and IA predictions were the key features of CIMM that drove the model output, and they exhibited stable monotonic positive relationships with mixture toxicity. The applicability domain assessment showed that in-domain samples achieved an external validation R2 of 0.768 which was markedly higher than 0.352 for out-of-domain samples, with lower prediction uncertainty for in-domain data. This study provides a high-accuracy, interpretable and applicability-domain-constrained framework for predicting the combined toxicity of complex mixture pollutants in aquatic environment, regardless of whether toxicity interactions occur within the mixtures. (Copyright © 2026 Elsevier B.V. All rights reserved.) |
| Competing Interests: | Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. |
| Contributed Indexing: | Keywords: CA model; Chlorella pyrenoidosa; IA model; Machine learning; Mixture toxicity |
| Substance Nomenclature: | 0 (Water Pollutants, Chemical) 0 (Complex Mixtures) |
| Entry Date(s): | Date Created: 20260806 Date Completed: 20260829 Latest Revision: 20260829 |
| Update Code: | 20260830 |
| DOI: | 10.1016/j.aquatox.2026.107967 |
| PMID: | 42561467 |
| Βάση Δεδομένων: | MEDLINE |
| FullText | Text: Availability: 0 CustomLinks: – Url: https://www.doi.org/10.1016/j.aquatox.2026.107967? Name: ScienceDirect (all content) (s7799221) Category: fullText Text: View record from ScienceDirect MouseOverText: View record from ScienceDirect – Url: https://resolver.ebsco.com/c/fiv2js/result?sid=EBSCO:cmedm&genre=article&issn=18791514&ISBN=&volume=299&issue=&date=20261001&spage=107967&pages=107967&title=Aquatic toxicology (Amsterdam, Netherlands)&atitle=Predicting%20the%20toxicity%20of%20multicomponent%20complex%20mixtures%20using%20a%20multi-feature%20fusion-based%20machine%20learning%20model.&aulast=Zhao%20Y&id=DOI:10.1016/j.aquatox.2026.107967 Name: Full Text Finder (for New FTF UI) (ns324271) Category: fullText Text: Full Text Finder MouseOverText: Full Text Finder |
|---|---|
| Header | DbId: cmedm DbLabel: MEDLINE An: 42561467 AccessLevel: 3 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 0 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Predicting the toxicity of multicomponent complex mixtures using a multi-feature fusion-based machine learning model. – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AU" term="%22Zhao+Y%22">Zhao Y</searchLink>; Key Laboratory of Water Pollution Control and Wastewater Resource of Anhui Province, College of Environment and Energy Engineering, Anhui Jianzhu University, Hefei, China; Anhui Gaodi Technology Co., LTD, Luan, China.<br /><searchLink fieldCode="AU" term="%22Jia+R%22">Jia R</searchLink>; Anhui Gaodi Technology Co., LTD, Luan, China.<br /><searchLink fieldCode="AU" term="%22Zhang+J%22">Zhang J</searchLink>; Key Laboratory of Water Pollution Control and Wastewater Resource of Anhui Province, College of Environment and Energy Engineering, Anhui Jianzhu University, Hefei, China; Anhui Gaodi Technology Co., LTD, Luan, China.<br /><searchLink fieldCode="AU" term="%22Xie+M%22">Xie M</searchLink>; Key Laboratory of Water Pollution Control and Wastewater Resource of Anhui Province, College of Environment and Energy Engineering, Anhui Jianzhu University, Hefei, China.<br /><searchLink fieldCode="AU" term="%22Xie+F%22">Xie F</searchLink>; Key Laboratory of Water Pollution Control and Wastewater Resource of Anhui Province, College of Environment and Energy Engineering, Anhui Jianzhu University, Hefei, China.<br /><searchLink fieldCode="AU" term="%22Mao+W%22">Mao W</searchLink>; Key Laboratory of Water Pollution Control and Wastewater Resource of Anhui Province, College of Environment and Energy Engineering, Anhui Jianzhu University, Hefei, China.<br /><searchLink fieldCode="AU" term="%22Zhang+J%22">Zhang J</searchLink>; Key Laboratory of Water Pollution Control and Wastewater Resource of Anhui Province, College of Environment and Energy Engineering, Anhui Jianzhu University, Hefei, China. Electronic address: ginnzy@ahjzu.edu.cn. – Name: TitleSource Label: Source Group: Src Data: <searchLink fieldCode="JN" term="%228500246%22">Aquatic toxicology (Amsterdam, Netherlands)</searchLink> [Aquat Toxicol] 2026 Oct; Vol. 299, pp. 107967. <i>Date of Electronic Publication: </i>2026 Aug 03. – Name: TypePub Label: Publication Type Group: TypPub Data: Journal Article – Name: Language Label: Language Group: Lang Data: English – Name: TitleSource Label: Journal Info Group: Src Data: <i>Publisher: </i><searchLink fieldCode="PB" term="%22Elsevier%2FNorth+Holland+Biomedical+Press%22">Elsevier/North Holland Biomedical Press </searchLink><i>Country of Publication: </i>Netherlands <i>NLM ID: </i>8500246 <i>Publication Model: </i>Print-Electronic <i>Cited Medium: </i>Internet <i>ISSN: </i>1879-1514 (Electronic) <i>Linking ISSN: </i><searchLink fieldCode="IS" term="%220166445X%22">0166445X </searchLink><i>NLM ISO Abbreviation: </i>Aquat Toxicol <i>Subsets: </i>MEDLINE – Name: PublisherInfo Label: Imprint Name(s) Group: PubInfo Data: <i>Original Publication</i>: Amsterdam, Netherlands : Elsevier/North Holland Biomedical Press, c1981- – Name: SubjectMESH Label: MeSH Terms Group: Su Data: <searchLink fieldCode="MM" term="%22Chlorella%22">Chlorella*</searchLink>/<searchLink fieldCode="MM" term="%22Chlorella+drug+effects%22">drug effects</searchLink> <br /><searchLink fieldCode="MM" term="%22Water+Pollutants%2C+Chemical%22">Water Pollutants, Chemical*</searchLink>/<searchLink fieldCode="MM" term="%22Water+Pollutants%2C+Chemical+toxicity%22">toxicity</searchLink> <br /><searchLink fieldCode="MM" term="%22Complex+Mixtures%22">Complex Mixtures*</searchLink>/<searchLink fieldCode="MM" term="%22Complex+Mixtures+toxicity%22">toxicity</searchLink> <br /><searchLink fieldCode="MM" term="%22Machine+Learning%22">Machine Learning*</searchLink><br /><searchLink fieldCode="MH" term="%22Predictive+Learning+Models%22">Predictive Learning Models</searchLink> ; <searchLink fieldCode="MH" term="%22Prediction+Algorithms%22">Prediction Algorithms</searchLink> – Name: Abstract Label: Abstract Group: Ab Data: Complex chemical mixtures are ubiquitous in aquatic environments posing substantial risks to aquatic organisms and human health, and the prediction of their combined toxicity is the key measure to control their ecological risks. This, however, still remains a major challenge for the two widely used classic standard additive models, concentration addition (CA) and independent action (IA), due to complex interactions, antagonism or synergism, within mixtures. Therefore, to conquer the dilemma, a novel interpretable machine learning model was constructed by integrating the predicted values by CA and IA with molecular descriptors (CIMM) against a freshwater organism Chlorella pyrenoidosa to predict the combined toxicity of complex mixture pollutants. The CIMM model stability was assessed by repeated modeling with ten random seeds, and its interpretability was examined by using mutual information, partial dependence plots and SHAP values. A dual-metric applicability domain was established based on kernel density estimation and k-nearest neighbor distance, and generalization performance was evaluated on a completely independent external validation set. The results showed that the CIMM framework achieved a mean test-set R<superscript>2</superscript> of 0.9080 ± 0.0127 across ten random-seed runs, which was implemented using the seed-specific optimal Random Forest or XGBoost algorithm. Compared to the results predicted by CA alone, CIMM increased R<superscript>2</superscript> by 32.2% and the RMSE decreased by 46.4%, while to those predicted by model using only molecular descriptors, the R<superscript>2</superscript> improved by approximately 8.87% and the RMSE decreased by 25.38%. CA and IA predictions were the key features of CIMM that drove the model output, and they exhibited stable monotonic positive relationships with mixture toxicity. The applicability domain assessment showed that in-domain samples achieved an external validation R<superscript>2</superscript> of 0.768 which was markedly higher than 0.352 for out-of-domain samples, with lower prediction uncertainty for in-domain data. This study provides a high-accuracy, interpretable and applicability-domain-constrained framework for predicting the combined toxicity of complex mixture pollutants in aquatic environment, regardless of whether toxicity interactions occur within the mixtures.<br /> (Copyright © 2026 Elsevier B.V. All rights reserved.) – Name: Abstract Label: Competing Interests Group: Ab Data: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. – Name: SubjectMinor Label: Contributed Indexing Group: Data: <i>Keywords: </i>CA model; Chlorella pyrenoidosa; IA model; Machine learning; Mixture toxicity – Name: NumberCAS Label: Substance Nomenclature Group: ID Data: 0 (Water Pollutants, Chemical)<br />0 (Complex Mixtures) – Name: DateEntry Label: Entry Date(s) Group: Date Data: <i>Date Created: </i>20260806 <i>Date Completed: </i>20260829 <i>Latest Revision: </i>20260829 – Name: DateUpdate Label: Update Code Group: Date Data: 20260830 – Name: DOI Label: DOI Group: ID Data: 10.1016/j.aquatox.2026.107967 – Name: AN Label: PMID Group: ID Data: 42561467 |
| PLink | https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=cmedm&AN=42561467 |
| RecordInfo | BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.1016/j.aquatox.2026.107967 Languages: – Code: eng Text: English PhysicalDescription: Pagination: StartPage: 107967 Subjects: – SubjectFull: Predictive Learning Models Type: general – SubjectFull: Prediction Algorithms Type: general – SubjectFull: Chlorella drug effects Type: general – SubjectFull: Water Pollutants, Chemical toxicity Type: general – SubjectFull: Complex Mixtures toxicity Type: general – SubjectFull: Machine Learning Type: general Titles: – TitleFull: Predicting the toxicity of multicomponent complex mixtures using a multi-feature fusion-based machine learning model. Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Zhao Y – PersonEntity: Name: NameFull: Jia R – PersonEntity: Name: NameFull: Zhang J – PersonEntity: Name: NameFull: Xie M – PersonEntity: Name: NameFull: Xie F – PersonEntity: Name: NameFull: Mao W – PersonEntity: Name: NameFull: Zhang J IsPartOfRelationships: – BibEntity: Dates: – D: 01 M: 10 Text: 2026 Oct Type: published Y: 2026 Identifiers: – Type: issn-electronic Value: 1879-1514 Numbering: – Type: volume Value: 299 Titles: – TitleFull: Aquatic toxicology (Amsterdam, Netherlands) Type: main |
| ResultId | 1 |