Logo PTI Logo FedCSIS

Position Papers of the 20th Conference on Computer Science and Intelligence Systems

Annals of Computer Science and Information Systems, Volume 44

Multitask Learning for Six-Pack Toxicity Prediction

, , ,

DOI: http://dx.doi.org/10.15439/2025F1171

Citation: Position Papers of the 20th Conference on Computer Science and Intelligence Systems, M. Bolanowski, M. Ganzha, L. Maciaszek, M. Paprzycki, D. Ślęzak (eds). ACSIS, Vol. 44, pages 93–97 ()

Full text

Abstract. The assessment of the six-pack toxicity, the crucial six systems and organ toxicities, is vital for ensuring the safe use of chemicals. Computational models capable of providing reliable predictions are acceptable for regulatory use to replace animal testing. However, data scarcity issues hindered the development of prediction models. This study proposed the first application of multitask learning to the six-pack toxicity for addressing data scarcity issues. Five algorithms were implemented and compared. Results showed that the distinct chemical space of tasks impedes the learning of shared representation of conventional algorithms, with performance worse than baseline models. In contrast, the MTForestNet algorithm built on a biological readacross concept performed best, with 3.1\% and 3.3\% improvement on AUC and accuracy, respectively. These findings demonstrate that biologically informed multitask learning can effectively overcome data scarcity and enhance toxicity prediction.

References

  1. J. V. B. Borba et al., “STopTox: An in Silico Alternative to Animal Testing for Acute Systemic and Topical Toxicity,” Environ. Health Perspect., vol. 130, no. 2, p. 27012, Feb. 2022, https://dx.doi.org/10.1289/EHP9341.
  2. Y. N. Fuadah, M. A. Pramudito, L. Firdaus, F. J. Vanheusden, and K. M. Lim, “QSAR Classification Modeling Using Machine Learning with a Consensus-Based Approach for Multivariate Chemical Hazard End Points,” ACS Omega, vol. 9, no. 51, pp. 50796–50808, Dec. 2024, https://dx.doi.org/10.1021/acsomega.4c09356.
  3. Y. Chushak, J. M. Gearhart, and R. A. Clewell, “Structural alerts and Machine learning modeling of ‘Six-pack’ toxicity as alternative to animal testing,” Comput. Toxicol., vol. 27, p. 100280, Aug. 2023, https://dx.doi.org/10.1016/j.comtox.2023.100280.
  4. B. Ramsundar et al., “Is Multitask Deep Learning Practical for Pharma?,” J. Chem. Inf. Model., vol. 57, no. 8, pp. 2068–2076, Aug. 2017, https://dx.doi.org/10.1021/acs.jcim.7b00146.
  5. N. Erickson et al., “AutoGluon-Tabular: Robust and Accurate AutoML for Structured Data,” Mar. 13, 2020, https://arxiv.org/abs/ arXiv:2003.06505. https://dx.doi.org/10.48550/arXiv.2003.06505.
  6. Z. Tan, Y. Li, W. Shi, and S. Yang, “A Multitask Approach to Learn Molecular Properties,” J. Chem. Inf. Model., vol. 61, no. 8, pp. 3824–3834, Aug. 2021, https://dx.doi.org/10.1021/acs.jcim.1c00646.
  7. X. Qian et al., “An Interpretable Multitask Framework BiLAT Enables Accurate Prediction of Cyclin-Dependent Protein Kinase Inhibitors,” J. Chem. Inf. Model., vol. 63, no. 11, pp. 3350–3368, Jun. 2023, https://dx.doi.org/10.1021/acs.jcim.3c00473.
  8. Y. Yuan Li et al., “Co-model for chemical toxicity prediction based on multi-task deep learning,” Mol. Inform., vol. 42, no. 5, p. e2200257, May 2023, https://dx.doi.org/10.1002/minf.202200257.
  9. X. Lin, Z. Quan, Z.-J. Wang, H. Huang, and X. Zeng, “A novel molecular representation with BiGRU neural networks for learning atom,” Brief. Bioinform., vol. 21, no. 6, pp. 2099–2111, Dec. 2020, https://dx.doi.org/10.1093/bib/bbz125.
  10. Y. Wang et al., “Multitask CapsNet: An Imbalanced Data Deep Learning Method for Predicting Toxicants,” ACS Omega, vol. 6, no. 40, pp. 26545–26555, Oct. 2021, https://dx.doi.org/10.1021/acsomega.1c03842.
  11. R.-H. Lin, P. Lin, C.-C. Wang, and C.-W. Tung, “A novel multitask learning algorithm for tasks with distinct chemical space: zebrafish toxicity prediction as an example,” J. Cheminformatics, vol. 16, no. 1, p. 91, Aug. 2024, https://dx.doi.org/10.1186/s13321-024-00891-4.
  12. L. Breiman, “Random Forests,” Mach. Learn., vol. 45, no. 1, pp. 5–32, Oct. 2001, https://dx.doi.org/10.1023/A:1010933404324.
  13. C.-C. Wang et al., “Using random forest to predict antimicrobial minimum inhibitory concentrations of nontyphoidal Salmonella in Taiwan,” Vet. Res., vol. 54, no. 1, p. 11, Feb. 2023, https://dx.doi.org/10.1186/s13567-023-01141-5.
  14. C.-Y. Chou, P. Lin, J. Kim, S.-S. Wang, C.-C. Wang, and C.-W. Tung, “Ensemble learning for predicting ex vivo human placental barrier permeability,” BMC Bioinformatics, vol. 22, no. Suppl 10, p. 629, Sep. 2022, https://dx.doi.org/10.1186/s12859-022-04937-y.
  15. C.-C. Wang, Y.-C. Liang, S.-S. Wang, P. Lin, and C.-W. Tung, “A machine learning-driven approach for prioritizing food contact chemicals of carcinogenic concern based on complementary in silico methods,” Food Chem. Toxicol. Int. J. Publ. Br. Ind. Biol. Res. Assoc., vol. 160, p. 112802, Feb. 2022, https://dx.doi.org/10.1016/j.fct.2021.112802.
  16. H.-L. Lin, Y.-W. Chiu, C.-C. Wang, and C.-W. Tung, “Computational prediction of Calu-3-based in vitro pulmonary permeability of chemicals,” Regul. Toxicol. Pharmacol. RTP, vol. 135, p. 105265, Nov. 2022, https://dx.doi.org/10.1016/j.yrtph.2022.105265.
  17. V. Svetnik, A. Liaw, C. Tong, J. C. Culberson, R. P. Sheridan, and B. P. Feuston, “Random forest: a classification and regression tool for compound classification and QSAR modeling,” J. Chem. Inf. Comput. Sci., vol. 43, no. 6, pp. 1947–1958, 2003, https://dx.doi.org/10.1021/ci034160g.
  18. C.-C. Wang, Y.-C. Lin, Y.-C. Lin, S.-R. Jhang, and C.-W. Tung, “Identification of informative features for predicting proinflammatory potentials of engine exhausts,” Biomed. Eng. Online, vol. 16, no. Suppl 1, p. 66, Aug. 2017, https://dx.doi.org/10.1186/s12938-017-0355-6.
  19. Y. Low et al., “Integrative chemical-biological read-across approach for chemical hazard classification,” Chem. Res. Toxicol., vol. 26, no. 8, pp. 1199–1208, Aug. 2013, https://dx.doi.org/10.1021/tx400110f.
  20. Y. Guo, L. Zhao, X. Zhang, and H. Zhu, “Using a hybrid read-across method to evaluate chemical toxicity based on chemical structure and biological data,” Ecotoxicol. Environ. Saf., vol. 178, pp. 178–187, Aug. 2019, https://dx.doi.org/10.1016/j.ecoenv.2019.04.019.