Derisking Future Agrochemicals before They Are Made: Large-Scale In Vitro Screening for In Silico Modeling of Thyroid Peroxidase Inhibition DOI

Martin Adamczewski,

Britta Nisius,

Nina Kausch‐Busies

и другие.

Chemical Research in Toxicology, Год журнала: 2024, Номер unknown

Опубликована: Сен. 20, 2024

Inhibition of thyroid peroxidase (TPO) is a known molecular initiating event for hormone dysregulation and toxicity. Consequently, TPO critical off-target the design safer agrochemicals. To date, fewer than 500 structurally characterized inhibitors are known, most comprehensive result set generated under identical conditions encompasses approximately 1000 compounds from subset ToxCast compound collection. Here we describe collaboration between wet lab data scientists combining large in vitro screen subsequent development an silico model predicting inhibition. The encompassed more 100,000 diverse drug-like agrochemical yielded 6000 novel inhibitors. On this foundation, applied different machine learning techniques compared their performance. We discuss use cases models research explain that recall particular importance when selecting virtual collections. Furthermore, show due to higher structural diversity our training data, final allowed better generalization trained on set. now have tool predict inhibition even molecules only available virtually, such as hits screenings, or consideration inclusion screening Structures activity 34,524 provided. This includes almost all inhibitors, including 3000 proprietary structures, proportion inactives.

Язык: Английский

Determination of the Zebrafish Embryo Developmental Toxicity Assessment (ZEDTA) as an Alternative Non-Mammalian Approach for the Safety Assessment of Agrochemicals. DOI Creative Commons
Jonathan S. Ball,

Anna Tochwin,

Matthew J. Winter

и другие.

Reproductive Toxicology, Год журнала: 2025, Номер unknown, С. 108837 - 108837

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

2

Threading the Eye of the Needle: Designing the Next Generation of Crop Protection Solutions in an Increasingly Deglobalized Regulatory Landscape DOI

Reza J. Rasoulpour,

Ashish Batra,

Guomin Shan

и другие.

Journal of Agricultural and Food Chemistry, Год журнала: 2025, Номер 73(3), С. 1801 - 1803

Опубликована: Янв. 22, 2025

The growing global needs for food and feed the impact of climate change on agriculture demand innovative sustainable solutions to protect against crop losses improve yields worldwide. To facilitate timely discovery development new protection products with favorable human health environmental safety profiles, approach methodologies (NAMs) is essential predict potential risk perform assessment early in process. Significant advances NAM research have been made recent years. We organized a Virtual Special Issue invited submissions seven high-quality articles reviews latest emerging NAMs predictive agrochemical development. Here we provide an introductory overview need innovation novel agricultural highlight findings from each article this special edition.

Язык: Английский

Процитировано

0

Cell Painting and Chemical Structure Read-Across Can Complement Each Other for Rat Acute Oral Toxicity Prediction in Chemical Early Derisking DOI

F. Camilleri,

J. Wenda,

C. Pecoraro-Mercier

и другие.

Chemical Research in Toxicology, Год журнала: 2024, Номер 37(11), С. 1851 - 1866

Опубликована: Ноя. 8, 2024

Early derisking decisions in the development of new chemical compounds enable identification novel candidates with improved safety profiles. In vivo studies are traditionally conducted early assessment acute oral toxicity crop protection products to avoid compounds, which considered "very acutely toxic", an lethal dose 50% (LD50) ≤ 60 mg/kg body weight. Those lengthy and costly raise ethical concerns, catalyzing use nonanimal alternatives. The objective our analysis was assess predictive efficacy read-across approaches for rats, comparing structure information, vitro biological data derived from Cell Painting profiling assay on U2OS cells, or combination both. Our findings indicate that classification as very toxic (LD50 mg/kg) not is possible using a approach, morphological profiles, When classifying structurally similar those training set, more (balanced accuracy 0.82). Conversely, when be classified were different profiles 0.72). Combining two models allowed set slightly improve predictions 0.85).

Язык: Английский

Процитировано

3

Cell Painting and chemical structure read-across can complement each other for rat acute oral toxicity prediction in chemical early de-risking DOI Creative Commons

F. Camilleri,

J. Wenda,

C. Pecoraro-Mercier

и другие.

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Апрель 28, 2024

ABSTRACT Early de-risking decisions in the development of new chemical compounds, enable identification novel candidates with improved safety profiles. In vivo studies are traditionally conducted early assessment acute oral toxicity crop protection products to avoid compounds which considered “very acutely toxic”, an Lethal Dose 50% (LD50) ≤ 60 mg/kg bodyweight. Those lengthy, costly, and raise ethical concerns, catalyzing use non-animal alternatives. The objective our analysis was assess predictive efficacy read-across approaches for rats, comparing structure information, vitro biological data derived from Cell Painting profiling assay on U2OS cells or combination both. Our findings indicate that classification as very toxic (LD50 mg/kg) not is possible using a approach, morphological profiles, When classifying structurally similar those training set, more (balanced accuracy 0.82). Conversely, when be classified were different profiles 0.72). Combining two models allowed slightly improve predictions 0.85). Abstract Figure For Table Contents Only

Язык: Английский

Процитировано

1

Derisking Future Agrochemicals before They Are Made: Large-Scale In Vitro Screening for In Silico Modeling of Thyroid Peroxidase Inhibition DOI

Martin Adamczewski,

Britta Nisius,

Nina Kausch‐Busies

и другие.

Chemical Research in Toxicology, Год журнала: 2024, Номер unknown

Опубликована: Сен. 20, 2024

Inhibition of thyroid peroxidase (TPO) is a known molecular initiating event for hormone dysregulation and toxicity. Consequently, TPO critical off-target the design safer agrochemicals. To date, fewer than 500 structurally characterized inhibitors are known, most comprehensive result set generated under identical conditions encompasses approximately 1000 compounds from subset ToxCast compound collection. Here we describe collaboration between wet lab data scientists combining large in vitro screen subsequent development an silico model predicting inhibition. The encompassed more 100,000 diverse drug-like agrochemical yielded 6000 novel inhibitors. On this foundation, applied different machine learning techniques compared their performance. We discuss use cases models research explain that recall particular importance when selecting virtual collections. Furthermore, show due to higher structural diversity our training data, final allowed better generalization trained on set. now have tool predict inhibition even molecules only available virtually, such as hits screenings, or consideration inclusion screening Structures activity 34,524 provided. This includes almost all inhibitors, including 3000 proprietary structures, proportion inactives.

Язык: Английский

Процитировано

0