top of page
Writer's pictureElena Fioravanzo

Enhancing Toxicity Prediction: A Deep Dive into Applicability Domains for Read-Across in Computational Toxicology


In the ever-evolving field of computational toxicology, the use of read-across approaches to predict the toxicity of chemicals has gained significant traction. This method, which allows us to infer the toxicity of a compound with little or no data from a similar compound with well-established data, is both promising and complex. A recent paper titled "A Strategy to Define Applicability Domains for Read-Across," authored by Cynthia Pestana et al., offers an in-depth exploration of a critical aspect of this approach—the definition of applicability domains (ADs).


I highly recommend diving into this paper to enrich your understanding of read-across and its applicability domains. It’s an essential read for anyone involved in toxicity prediction, particularly in regulatory settings. By familiarizing yourself with these strategies, you’ll be better equipped to justify your predictions and contribute more effectively to the field of toxicology.


For those looking to deepen their understanding of read-across, I highly recommend checking out our course NAMs - Use and application of QSAR and read-across, which provides valuable insights that complement the discussion in this post.


Why This Paper is Essential for Toxicologists:

As toxicologists, we often face challenges when justifying read-across predictions, especially for regulatory purposes. The primary hurdle lies in determining whether the properties of an analogue or a group of chemicals fall within a domain where a reliable read-across prediction can be made. This paper presents a robust, comprehensive, and flexible strategy that addresses these challenges head-on.


Key Takeaways:


  1. Applicability Domains Defined: The paper provides a clear framework for defining applicability domains in the context of read-across. It synthesizes existing knowledge into a practical strategy that considers chemical structure, toxicodynamics, and toxicokinetics.


  2. Adaptability Across Scenarios: One of the most valuable aspects of this strategy is its adaptability. Whether you're working with analogues or categories, or even dealing with scenarios involving common metabolites or biological profiles, this strategy is designed to accommodate various situations.


  3. Practical Implementation: The authors demonstrate how this strategy can be implemented using real-world examples, such as the repeated dose toxicity of triazoles and imidazoles. This practical approach makes it easier for toxicologists to apply the concepts in their own work.


  4. Addressing the Uncertainties: The strategy also addresses the perennial question of what constitutes an "acceptable" degree of similarity, offering guidance on how to navigate this grey area.


Why You Should Read This Paper:

For toxicologists who are keen to expand their knowledge, this paper is a goldmine. It not only provides a deep understanding of the challenges associated with read-across but also equips you with the tools needed to enhance the reliability of your predictions. By adopting the strategies outlined in this paper, you can improve the accuracy of your toxicity predictions, making your work more impactful and scientifically robust.


Cynthia Pestana, Steven J. Enoch, James W. Firman, Judith C. Madden, Nicoleta Spînu, Mark T.D. Cronin, A strategy to define applicability domains for read-across, Computational Toxicology, Volume 22, 2022, 100220, ISSN 2468-1113,

133 views0 comments

Comments


bottom of page