Xueshu Li, Joe J Lim, Cayen Rong, Hans-Joachim Lehmler, Julia Yue Cui
{"title":"Deconjugation of Polychlorinated Biphenyl Sulfates to Hydroxylated PCBs by Anaerobically Cultured Mouse and Human Gut Microbiota.","authors":"Xueshu Li, Joe J Lim, Cayen Rong, Hans-Joachim Lehmler, Julia Yue Cui","doi":"10.1021/acs.chemrestox.5c00016","DOIUrl":"10.1021/acs.chemrestox.5c00016","url":null,"abstract":"<p><p>The role of the gut microbiome in metabolizing polychlorinated biphenyls (PCBs), toxic environmental contaminants, and their metabolites remains unclear. This study used mouse and human microbiomes in anaerobic cultures to investigate the metabolism of PCB sulfate to hydroxylated PCBs (OH-PCBs). All microbiomes enzymatically hydrolyzed PCB sulfates. Higher chlorinated PCB sulfates were metabolized more readily. Male mouse microbiomes exhibited more PCB sulfate hydrolysis to OH-PCBs than female mouse microbiomes. Human microbiomes metabolized PCB sulfates to a more considerable extent than mouse microbiomes. They also showed variability in PCB sulfate metabolism, depending on the microbial communities. These findings suggest that the microbiome contributes to PCB metabolism.</p>","PeriodicalId":31,"journal":{"name":"Chemical Research in Toxicology","volume":" ","pages":"557-560"},"PeriodicalIF":3.7,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12015953/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143699092","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Caroline Andolfato Sanchez, Estefani Maria Treviso, Cecília Cristina de Souza Rocha, Lusânia Maria Greggi Antunes
{"title":"Diallyl Disulfide Reduces Ethyl Carbamate-Induced Cytotoxicity and Apoptosis in Intestinal and Hepatic Cells.","authors":"Caroline Andolfato Sanchez, Estefani Maria Treviso, Cecília Cristina de Souza Rocha, Lusânia Maria Greggi Antunes","doi":"10.1021/acs.chemrestox.4c00439","DOIUrl":"10.1021/acs.chemrestox.4c00439","url":null,"abstract":"<p><p>Epidemiological studies indicate that lifestyle and dietary habits are associated with an increasing cancer incidence. Consuming fermented foods and alcoholic beverages and smoking can expose humans to ethyl carbamate (EC), a probable human carcinogen classified as group 2A by the International Agency for Research on Cancer (IARC). Increasing the intake of bioactive compounds can reduce EC-induced toxicity. Diallyl disulfide (DADS), found in garlic, may protect against damage induced by chemical agents and natural compounds. Here, the potential protective effect of DADS against EC was investigated by evaluating EC-induced cytotoxicity, DNA damage, apoptosis, and reactive oxygen species production in colorectal adenocarcinoma (Caco-2) and hepatocarcinoma (HepG2) cells. To this end, resazurin, comet, and annexin V-FITC staining assays and CM-H<sub>2</sub>DCFDA markers were used to evaluate the effect on Caco-2 and HepG2 cells of protocols combining DADS (10-120 μM) and EC (80 mM). The protocols were as follows: (i) cells pretreated with DADS for 2 h and exposed to EC for 24 h; (ii) cells pretreated with DADS for 24 h and exposed to EC for 24 h; (iii) cells simultaneously exposed to DADS and EC for 24 h; (iv) cells exposed to EC for 24 h and treated with DADS for 2 h. EC induced cytotoxicity and apoptosis in Caco-2 and HepG2 cells and oxidative damage in Caco-2 cells. Combined exposure to DADS and EC for 24 h decreased EC-mediated cytotoxicity and apoptosis in both Caco-2 and HepG2 cells. These findings encourage further studies on the mechanisms of action of the combined DADS and EC.</p>","PeriodicalId":31,"journal":{"name":"Chemical Research in Toxicology","volume":" ","pages":"623-634"},"PeriodicalIF":3.7,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12015961/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143717600","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"RNA Methylation and Transcriptome Analysis Reveal Key Regulatory Pathways Related to Cadmium-Induced Liver Damage.","authors":"Hao Huang, Guoliang Li, Sihui Guo, Kaile Li, Wei Li, Qinwen Zhou, Zhini He, Xingfen Yang, Lili Liu, Qinzhi Wei","doi":"10.1021/acs.chemrestox.4c00539","DOIUrl":"10.1021/acs.chemrestox.4c00539","url":null,"abstract":"<p><p>Cadmium (Cd) is a prevalent environmental and industrial contaminant that causes significant damage to liver function. However, the role of m<sup>6</sup>A methylation─a critical epigenetic modification─in Cd-induced liver injury remains poorly understood. This study aimed to investigate the effects of m<sup>6</sup>A methylation in Cd-induced liver damage. A mouse model of Cd-induced liver injury was established, and exposure to CdCl<sub>2</sub> (20 mg/kg) for 90 days resulted in reduced m<sup>6</sup>A methylation levels. Using methylated RNA immunoprecipitation sequencing (MeRIP-seq) and RNA sequencing (RNA-Seq), we characterized the m<sup>6</sup>A methylation profiles in both control and Cd-exposed groups. A total of 8355 unique m<sup>6</sup>A peaks and 1,101 unique m<sup>6</sup>A-modified genes were identified. Among these, 673 genes exhibited differential m<sup>6</sup>A methylated modifications, including 463 hyper-methylated and 210 hypo-methylated genes. Conjoint analysis of MeRIP-seq and RNA-Seq data unveiled genes that showed both differential methylation and expression. These genes were significantly enriched in the AGE-RAGE and PI3K-Akt signaling pathway. Through bioinformatics screening, five key genes (<i>Il-1β</i>, <i>Ccl2</i>, <i>Tlr2</i>, <i>Itgax</i>, and <i>Ccr2</i>) were identified, and expression validation indicated that <i>Itgax</i> and <i>Ccr2</i> may play pivotal roles in Cd-induced liver injury. Notably, elevated expression of methyltransferase-like 14 (METTL14) was observed in both in vivo and in vitro models. Inhibition of <i>Mettl14</i> can regulate Cd-induced liver inflammation through m<sup>6</sup>A-dependent regulation of <i>Ccr2</i> expression. Collectively, our findings highlight the crucial role of Mettl14 and Ccr2 in Cd-induced liver injury, providing novel insights into the epigenetic mechanisms underlying liver diseases and potential biomarkers for diagnosis and therapy.</p>","PeriodicalId":31,"journal":{"name":"Chemical Research in Toxicology","volume":" ","pages":"717-732"},"PeriodicalIF":3.7,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143707734","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Marina Garcia de Lomana, Domenico Gadaleta, Marian Raschke, Robert Fricke, Floriane Montanari
{"title":"Predicting Liver-Related In Vitro Endpoints with Machine Learning to Support Early Detection of Drug-Induced Liver Injury.","authors":"Marina Garcia de Lomana, Domenico Gadaleta, Marian Raschke, Robert Fricke, Floriane Montanari","doi":"10.1021/acs.chemrestox.4c00453","DOIUrl":"10.1021/acs.chemrestox.4c00453","url":null,"abstract":"<p><p>Drug-induced liver injury (DILI) is a major cause of drug development failures and postmarket drug withdrawals, posing significant challenges to public health and pharmaceutical research. The biological mechanisms leading to DILI are highly complex and the adverse reaction is often difficult to foresee. Hence, mechanistic insights into DILI, as well as machine learning models to predict molecular events that trigger adverse outcomes, pharmacokinetics and pharmacodynamics in the liver, are essential tools for understanding and preventing DILI. In this study, we collected a comprehensive data set of 28 in vitro endpoints related to liver toxicity and function, as well as data specific to DILI, to explore the potential of multi-task learning for their prediction. We demonstrate the benefits of ensemble modeling and provide an uncertainty estimation based on the standard deviation of the predictions to define an applicability domain for the models. Available assays at Bayer for two of the endpoints (Bile salt export pump (BSEP) inhibition and phospholipidosis) were run on a set of public compounds and used for further evaluation (data provided in the Supporting Information). Additionally, we conducted an in-depth data analysis of the relationships among the different endpoints, as well as with DILI. The presented models can be used to derive a \"Virtual Liver Safety Profile\" showcasing the predicted activity of a compound on the selected endpoints to support the prioritization of assays and the elucidation of modes of action.</p>","PeriodicalId":31,"journal":{"name":"Chemical Research in Toxicology","volume":" ","pages":"656-671"},"PeriodicalIF":3.7,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12015958/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143595774","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Molecular Insights into Propofol's Neurotoxic Effects: Targeting the HTR1A/cAMP Signaling Pathway.","authors":"Gongrui Zhou, Shubin Zheng, Yuhai Xu","doi":"10.1021/acs.chemrestox.4c00339","DOIUrl":"10.1021/acs.chemrestox.4c00339","url":null,"abstract":"<p><p>Propofol, a commonly used anesthetic in clinical practice, is favored for its rapid onset and short duration of action. Despite its widespread use, the potential neurotoxic effects of propofol remain insufficiently understood. This study utilized high-throughput transcriptome sequencing and network pharmacology to investigate the mechanisms by which propofol induces neurotoxicity in rat hippocampal neural progenitor cells (NPCs), focusing on the HTR1A/cAMP signaling pathway. Our findings reveal that propofol significantly inhibits the HTR1A/cAMP pathway, leading to altered expression of key genes that affect neuronal activity, inflammatory responses, and apoptosis. In vivo experiments further demonstrate that propofol impairs spatial learning and memory in rats, an effect that is partially reversed by overexpression of HTR1A. These results not only elucidate the molecular mechanisms underlying propofol-induced neuronal damage but also provide critical insights into the safe application of propofol in clinical settings.</p>","PeriodicalId":31,"journal":{"name":"Chemical Research in Toxicology","volume":" ","pages":"561-572"},"PeriodicalIF":3.7,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143750179","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sandra I Salido, Matthew Hassink, Kenneth Taylor, Elizabeth A Cowan, Morgann S Woods, Shanga Lee, Oleksii Motorykin, Bryan Hearn, Liza Valentín-Blasini, Clifford H Watson
{"title":"Benzo[a]pyrene in the Mainstream Smoke of Little Cigars.","authors":"Sandra I Salido, Matthew Hassink, Kenneth Taylor, Elizabeth A Cowan, Morgann S Woods, Shanga Lee, Oleksii Motorykin, Bryan Hearn, Liza Valentín-Blasini, Clifford H Watson","doi":"10.1021/acs.chemrestox.4c00411","DOIUrl":"10.1021/acs.chemrestox.4c00411","url":null,"abstract":"<p><p>Little cigars are similar to cigarettes, with respect to dimensions, filters, and overall appearance. Some smokers also use little cigars as substitutes for cigarettes. Comparison of little cigars with cigarettes is relevant to understanding their respective public health impact. To understand their relative toxicities, mainstream smoke yields of benzo[a]pyrene (B[a]P), a human carcinogen, were measured for 60 commercial little cigars. The little cigars were smoked on a linear smoking machine using the International Organization of Standardization (ISO) nonintense and Canadian Intense (CI) smoking regimens followed by analysis with a validated gas chromatography/mass spectrometry (GC/MS) method. The average analytical quantitative variability of the measured little cigar constituents was lower compared to previously tested commercial cigarettes (%RSD 9.6 vs 14.5, respectively). B[a]P yields ranged from 14.5-44.0 ng/cigar (ISO) and 24.0-65.7 ng/cigar (CI). The mean ISO yield is 25.5 ng/cigar versus the CI yield of 42.2 ng/cigar, which are 2.5- and 2-fold greater, respectively, than the corresponding mean cigarette yields. When normalized to tobacco weight, B[a]P yields of the little cigars are 1.5- (ISO) and 1.3- (CI) fold greater than cigarette yields. B[a]P smoke yields are known to correlate with tobacco weight. The little cigar B[a]P yield correlations to tobacco weight (CI <i>R</i><sup>2</sup> = 0.35; ISO <i>R</i><sup>2</sup> = 0.24) are similar to cigarette yield correlations (CI <i>R</i><sup>2</sup> = 0.31; ISO <i>R</i><sup>2</sup> = 0.21). Other physical properties (i.e., filter length, filter ventilation, and packing density) that may impact B[a]P smoke yields for the little cigars had very weak correlations. Except for cigarette packing density, cigars and cigarettes have similar correlations between B[a]P yields and physical design parameters. In summary, the little cigars, although physically similar to cigarettes, differ in smoke chemistry by generating higher B[a]P yields, even when normalized to tobacco weight.</p>","PeriodicalId":31,"journal":{"name":"Chemical Research in Toxicology","volume":" ","pages":"609-622"},"PeriodicalIF":3.7,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12011531/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143583937","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Kevin J Murray, Dylan Mckeon, Chiara Lecchi, Laura Maertens, Peter W Villalta, Silvia Balbo
{"title":"Positive Ion Tandem Mass Spectrometry Offers Enhanced Structural Insights for the Discovery of Mercapturic Acids.","authors":"Kevin J Murray, Dylan Mckeon, Chiara Lecchi, Laura Maertens, Peter W Villalta, Silvia Balbo","doi":"10.1021/acs.chemrestox.4c00446","DOIUrl":"10.1021/acs.chemrestox.4c00446","url":null,"abstract":"<p><p>Urinary mercapturic acids represent valuable biological markers of chemical exposure and detoxification mechanisms. Characterization of this class of compound has historically employed LC-MS/MS analytical platforms using negative ion mode. In this study, we report the first application of a UHPLC-MS/MS method using positive ion mode detection for the unbiased characterization of mercapturic acids. A preliminary spectral library of synthetically available mercapturic acids was generated to evaluate fragmentation pathways of mercapturic acids in positive mode. From our findings, we propose a discovery method that utilizes a neutral loss monitoring paradigm based on two diagnostic fragmentation pathways of mercapturic acids. Using a cohort of 20 nonsmokers and 20 smokers, we detected 180 putative mercapturic acid signatures that exhibited a high degree of reproducibility. Following a combination of multivariate and univariate statistics, we found 33 putative mercapturic acids associated with smoking status. The increased structural insights of analytical profiling in positive mode enable more informative annotation of discovery results. Using the latest structural prediction technology, we were able to assign preliminary structural identifications to these features and eliminate likely false positive detections. From our workflow, we discovered a previously unreported mercapturic acid with a strong association with tobacco cigarette usage and putatively identified it as <i>N</i>-acetyl-<i>S</i>-(2-ethyl-3-pyridine)-l-cysteine, a potential metabolite of nicotine pyrolysis product 3-ethenylpyridine.</p>","PeriodicalId":31,"journal":{"name":"Chemical Research in Toxicology","volume":" ","pages":"635-646"},"PeriodicalIF":3.7,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143595773","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Alicia Thiel, Sarah Heider, Kira Bieck, Vivien Michaelis, Tanja Schwerdtle, Franziska Ebert, Julia Bornhorst
{"title":"Genotoxicity Assessment of Co(II) and Ni(II) in HepG2 Cells: Insights into Combined Metal Exposure.","authors":"Alicia Thiel, Sarah Heider, Kira Bieck, Vivien Michaelis, Tanja Schwerdtle, Franziska Ebert, Julia Bornhorst","doi":"10.1021/acs.chemrestox.4c00518","DOIUrl":"10.1021/acs.chemrestox.4c00518","url":null,"abstract":"<p><p>The usage of cobalt (Co) and nickel (Ni) in numerous commercial, industrial, and military applications causes widespread exposure nowadays, and concerns are rising about adverse impacts on human health. Emphasis is on the respiratory system, with both metals classified as (possibly) carcinogenic upon inhalation by the International Agency for Research on Cancer (IARC), but limited data are available upon oral exposure. Therefore, this study aims to evaluate the <i>in vitro</i> genotoxicity of Co(II) and Ni(II) and their combination in HepG2 cells, since exposure of those environmental pollutants occurs realistically in concert. Here, Co(II) exposure led to the induction of single-strand breaks and oxidative DNA damage detected by the Comet assay as FPG-sensitive sites, while Ni(II) increased the abundance of γ-H2AX, an indicator for double-strand breaks. Notably, combined exposure to Co(II) and Ni(II) resulted in enhanced DNA damage, especially at the chromosomal level, with increased formation of micronuclei as well as polynucleated cells, indicating a stronger effect compared to single exposure. Furthermore, both metals induced the DNA damage response pathway PARylation. As this process involves the consumption of large amounts of cellular NAD<sup>+</sup> after DNA damage, the energy state was assessed upon exposure with Co(II) and Ni(II). Current data indicate that especially Co(II) altered the cellular energy state. This study reveals distinct mechanisms of DNA damage exhibited by Co(II) and Ni(II), which were enhanced after a combined treatment. This highlights the need for further research to estimate the genotoxic potential of targeting cells upon oral intake with increasing environmental entry.</p>","PeriodicalId":31,"journal":{"name":"Chemical Research in Toxicology","volume":" ","pages":"695-704"},"PeriodicalIF":3.7,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143802006","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Combining Molecular Docking and Pharmacophore Models Predicts Ligand Binding of Endocrine-Disrupting Chemicals to Nuclear Receptors.","authors":"Asma Sellami, Matthieu Montes, Nathalie Lagarde","doi":"10.1021/acs.chemrestox.4c00455","DOIUrl":"10.1021/acs.chemrestox.4c00455","url":null,"abstract":"<p><p>Nuclear receptors form a family of proteins capable of accommodating a wide variety of small molecules in their ligand binding domain, ranging from therapeutic compounds to endocrine-disrupting chemicals. The rapid identification of these compounds, especially within the latter category, is of paramount importance. Using data extracted from the CompTox Dashboard, an Environmental Protection Agency initiative, we assessed the effectiveness of a combination of molecular docking and pharmacophore models in identifying ligands binding to six nuclear receptors: androgen receptor, estrogen receptor alpha, estrogen receptor beta, glucocorticoid receptor, peroxisome proliferator-activated receptor gamma, and thyroid hormone receptor alpha. For each nuclear receptor, we selected a specifically designed and optimized in silico protocol that, in conjunction with experimental assays, can prioritize compounds for further evaluation to detect any potential toxicological concerns.</p>","PeriodicalId":31,"journal":{"name":"Chemical Research in Toxicology","volume":" ","pages":"672-685"},"PeriodicalIF":3.7,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143770725","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Aljoša Smajić, Thomas Steger-Hartmann, Gerhard. F. Ecker and Anke Hackl*,
{"title":"Data Exploration for Target Predictions Using Proprietary and Publicly Available Data Sets","authors":"Aljoša Smajić, Thomas Steger-Hartmann, Gerhard. F. Ecker and Anke Hackl*, ","doi":"10.1021/acs.chemrestox.4c0034710.1021/acs.chemrestox.4c00347","DOIUrl":"https://doi.org/10.1021/acs.chemrestox.4c00347https://doi.org/10.1021/acs.chemrestox.4c00347","url":null,"abstract":"<p >When applying machine learning (ML) approaches for the prediction of bioactivity, it is common to collect data from different assays or sources and combine them into single data sets. However, depending on the data domains and sources from which these data are retrieved, bioactivity data for the same macromolecular target may show a high variance of values (looking at a single compound) and cover very different parts of the chemical space as well as the bioactivity range (looking at the whole data set). The effectiveness and applicability domain of the resulting prediction models may be strongly influenced by the sources from which their training data were retrieved. Therefore, we investigated the chemical space and active/inactive distribution of proprietary pharmaceutical data from Bayer AG and the publicly available ChEMBL database, and their impact when applied as training data for classification models. For this end, we applied two different sets of descriptors in combination with different ML algorithms. The results show substantial differences in chemical space between the two different data sources, leading to suboptimal prediction performance when models are applied to domains other than their training data. MCC values between −0.34 and 0.37 among all targets were retrieved, indicating suboptimal model performance when models trained on Bayer AG data were tested on ChEMBL data and vice versa. The mean Tanimoto similarity of the nearest neighbors between these two data sources indicated similarities for 31 targets equal to or less than 0.3. Interestingly, all applied methods to assess overlap of chemical space of the two data sources to predict the applicability of models beyond their training data sets did not correlate with observed performances. Finally, we applied different strategies for creating mixed training data sets based on both public and proprietary sources, using assay format (cell-based and cell-free) information and Tanimoto similarities.</p>","PeriodicalId":31,"journal":{"name":"Chemical Research in Toxicology","volume":"38 5","pages":"820–833 820–833"},"PeriodicalIF":3.7,"publicationDate":"2025-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://pubs.acs.org/doi/epdf/10.1021/acs.chemrestox.4c00347","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144083477","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}