Humans are exposed to pesticides via diet, occupational exposures, and environmental sources. While individual pesticides are unlikely to cause adverse effects, exposure to multiple pesticides may still be harmful. This project proposes a tractable “lab-in-the-loop” approach to determine potential additive effects of pesticide mixtures.

While the toxicity of pesticides is often assessed one chemical at a time, in real-world situations, individuals are exposed to mixtures of pesticides that can act through the same molecular pathway. This means that pesticides could have additive effects on health and evaluating them as a group would more accurately assess these risks. However, testing all the possible combinations of pesticides for additive effects is logistically infeasible. In this project, a “lab-in-the-loop” approach will be used to estimate pesticide modes of action and potential effects in human cells using high-throughput screening methods across diverse cell types.

Using machine learning methods, the impact of compound mixtures will be determined with predictions validated through follow-up experiments. If successful, the approach would annotate pesticide affects across cellular targets, providing a quantitative framework to assess exposure risk for pesticide mixtures.