In SilicoPlatforms for Predictive Ecotoxicology
Yong Oh Lee, Baeckkyoung Sung · 2021
Data-driven computational predictions of chemical toxicity have drastically accelerated the toxicological screening process. Such in silico methods have been proven to reduce the duration and cost of environmental toxicity tests, offering high-precision prediction for humans and various ecological species. Specifically, recent developments in machine learning frameworks (including deep learning) have provided next-generation in silico tools for predictive ecotoxicology. The purpose of this chapter is to provide a brief introduction to in silico toxicological modeling methods based on machine learning and deep learning. A concise overview of their technological background is provided, as well as their applications to ecotoxicology. Finally, current challenges and future perspectives related to modeling toolkit developments are discussed.