A Machine Learning Model to Estimate Toxicokinetic Half-Lives of Per- and Polyfluoro-Alkyl Substances (PFAS) in Multiple Species
Per- and polyfluoroalkyl substances (PFAS) are a diverse group of man-made chemicals that are commonly found in body tissues. The toxicokinetics of most PFAS are currently uncharacterized, but long half-lives (<i>t</i><sub>½</sub>) have been observed in some cases. Knowledge...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-01-01
|
| Series: | Toxics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2305-6304/11/2/98 |
| _version_ | 1827755328167477248 |
|---|---|
| author | Daniel E. Dawson Christopher Lau Prachi Pradeep Risa R. Sayre Richard S. Judson Rogelio Tornero-Velez John F. Wambaugh |
| author_facet | Daniel E. Dawson Christopher Lau Prachi Pradeep Risa R. Sayre Richard S. Judson Rogelio Tornero-Velez John F. Wambaugh |
| author_sort | Daniel E. Dawson |
| collection | DOAJ |
| description | Per- and polyfluoroalkyl substances (PFAS) are a diverse group of man-made chemicals that are commonly found in body tissues. The toxicokinetics of most PFAS are currently uncharacterized, but long half-lives (<i>t</i><sub>½</sub>) have been observed in some cases. Knowledge of chemical-specific <i>t</i><sub>½</sub> is necessary for exposure reconstruction and extrapolation from toxicological studies. We used an ensemble machine learning method, random forest, to model the existing in vivo measured <i>t</i><sub>½</sub> across four species (human, monkey, rat, mouse) and eleven PFAS. Mechanistically motivated descriptors were examined, including two types of surrogates for renal transporters: (1) physiological descriptors, including kidney geometry, for renal transporter expression and (2) structural similarity of defluorinated PFAS to endogenous chemicals for transporter affinity. We developed a classification model for <i>t</i><sub>½</sub> (Bin 1: <12 h; Bin 2: <1 week; Bin 3: <2 months; Bin 4: >2 months). The model had an accuracy of 86.1% in contrast to 32.2% for a y-randomized null model. A total of 3890 compounds were within domain of the model, and <i>t</i><sub>½</sub> was predicted using the bin medians: 4.9 h, 2.2 days, 33 days, and 3.3 years. For human <i>t</i><sub>½</sub>, 56% of PFAS were classified in Bin 4, 7% were classified in Bin 3, and 37% were classified in Bin 2. This model synthesizes the limited available data to allow tentative extrapolation and prioritization. |
| first_indexed | 2024-03-11T08:03:45Z |
| format | Article |
| id | doaj.art-5b25ff94479248d18f78fa6271bf1581 |
| institution | Directory Open Access Journal |
| issn | 2305-6304 |
| language | English |
| last_indexed | 2024-03-11T08:03:45Z |
| publishDate | 2023-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Toxics |
| spelling | doaj.art-5b25ff94479248d18f78fa6271bf15812023-11-16T23:36:55ZengMDPI AGToxics2305-63042023-01-011129810.3390/toxics11020098A Machine Learning Model to Estimate Toxicokinetic Half-Lives of Per- and Polyfluoro-Alkyl Substances (PFAS) in Multiple SpeciesDaniel E. Dawson0Christopher Lau1Prachi Pradeep2Risa R. Sayre3Richard S. Judson4Rogelio Tornero-Velez5John F. Wambaugh6U.S. Environmental Protection Agency, Office of Research and Development, Center for Computational Toxicology and Exposure, 109 T.W. Alexander Drive, Research Triangle Park, NC 27711, USAU.S. Environmental Protection Agency, Office of Research and Development, Center for Public Health and Environmental Assessment, 109 T.W. Alexander Drive, Research Triangle Park, NC 277011, USAU.S. Environmental Protection Agency, Office of Research and Development, Center for Computational Toxicology and Exposure, 109 T.W. Alexander Drive, Research Triangle Park, NC 27711, USAU.S. Environmental Protection Agency, Office of Research and Development, Center for Computational Toxicology and Exposure, 109 T.W. Alexander Drive, Research Triangle Park, NC 27711, USAU.S. Environmental Protection Agency, Office of Research and Development, Center for Computational Toxicology and Exposure, 109 T.W. Alexander Drive, Research Triangle Park, NC 27711, USAU.S. Environmental Protection Agency, Office of Research and Development, Center for Computational Toxicology and Exposure, 109 T.W. Alexander Drive, Research Triangle Park, NC 27711, USAU.S. Environmental Protection Agency, Office of Research and Development, Center for Computational Toxicology and Exposure, 109 T.W. Alexander Drive, Research Triangle Park, NC 27711, USAPer- and polyfluoroalkyl substances (PFAS) are a diverse group of man-made chemicals that are commonly found in body tissues. The toxicokinetics of most PFAS are currently uncharacterized, but long half-lives (<i>t</i><sub>½</sub>) have been observed in some cases. Knowledge of chemical-specific <i>t</i><sub>½</sub> is necessary for exposure reconstruction and extrapolation from toxicological studies. We used an ensemble machine learning method, random forest, to model the existing in vivo measured <i>t</i><sub>½</sub> across four species (human, monkey, rat, mouse) and eleven PFAS. Mechanistically motivated descriptors were examined, including two types of surrogates for renal transporters: (1) physiological descriptors, including kidney geometry, for renal transporter expression and (2) structural similarity of defluorinated PFAS to endogenous chemicals for transporter affinity. We developed a classification model for <i>t</i><sub>½</sub> (Bin 1: <12 h; Bin 2: <1 week; Bin 3: <2 months; Bin 4: >2 months). The model had an accuracy of 86.1% in contrast to 32.2% for a y-randomized null model. A total of 3890 compounds were within domain of the model, and <i>t</i><sub>½</sub> was predicted using the bin medians: 4.9 h, 2.2 days, 33 days, and 3.3 years. For human <i>t</i><sub>½</sub>, 56% of PFAS were classified in Bin 4, 7% were classified in Bin 3, and 37% were classified in Bin 2. This model synthesizes the limited available data to allow tentative extrapolation and prioritization.https://www.mdpi.com/2305-6304/11/2/98perfluoro-alkyl substancesPFAShalf-lifemachine learning modeltoxicokinetics |
| spellingShingle | Daniel E. Dawson Christopher Lau Prachi Pradeep Risa R. Sayre Richard S. Judson Rogelio Tornero-Velez John F. Wambaugh A Machine Learning Model to Estimate Toxicokinetic Half-Lives of Per- and Polyfluoro-Alkyl Substances (PFAS) in Multiple Species Toxics perfluoro-alkyl substances PFAS half-life machine learning model toxicokinetics |
| title | A Machine Learning Model to Estimate Toxicokinetic Half-Lives of Per- and Polyfluoro-Alkyl Substances (PFAS) in Multiple Species |
| title_full | A Machine Learning Model to Estimate Toxicokinetic Half-Lives of Per- and Polyfluoro-Alkyl Substances (PFAS) in Multiple Species |
| title_fullStr | A Machine Learning Model to Estimate Toxicokinetic Half-Lives of Per- and Polyfluoro-Alkyl Substances (PFAS) in Multiple Species |
| title_full_unstemmed | A Machine Learning Model to Estimate Toxicokinetic Half-Lives of Per- and Polyfluoro-Alkyl Substances (PFAS) in Multiple Species |
| title_short | A Machine Learning Model to Estimate Toxicokinetic Half-Lives of Per- and Polyfluoro-Alkyl Substances (PFAS) in Multiple Species |
| title_sort | machine learning model to estimate toxicokinetic half lives of per and polyfluoro alkyl substances pfas in multiple species |
| topic | perfluoro-alkyl substances PFAS half-life machine learning model toxicokinetics |
| url | https://www.mdpi.com/2305-6304/11/2/98 |
| work_keys_str_mv | AT danieledawson amachinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT christopherlau amachinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT prachipradeep amachinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT risarsayre amachinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT richardsjudson amachinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT rogeliotornerovelez amachinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT johnfwambaugh amachinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT danieledawson machinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT christopherlau machinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT prachipradeep machinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT risarsayre machinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT richardsjudson machinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT rogeliotornerovelez machinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies AT johnfwambaugh machinelearningmodeltoestimatetoxicokinetichalflivesofperandpolyfluoroalkylsubstancespfasinmultiplespecies |