Functional non-parametric mixed effects models for cytotoxicity assessment and clustering

Area of Science:

• Environmental Science
• Toxicology
• Biostatistics

Background:

• Chemicals in the environment necessitate regulatory oversight.
• Traditional cytotoxicity research relies on point measurements like LD50.
• Understanding chemical impact requires analyzing time-dependent cellular responses.

Purpose of the Study:

• To develop and apply functional mixed effects models for analyzing time-dependent cytotoxicity data.
• To identify chemical modes of action based on cellular response curves.
• To improve the efficiency of cytotoxicity research and chemical safety assessments.

Main Methods:

• Application of functional mixed effects models to time-dependent cellular response curves.
• Utilizing functional principal components and B-splines for data analysis.
• Employing k-means and self-organizing maps for cluster analysis of identified curve features.

Main Results:

• Differences in cellular response curves were identified, correlating with chemical modes of action.
• Key curve features were extracted for effective cluster analysis.
• The proposed methodology offers a significant speed-up for future cytotoxicity studies.

Conclusions:

• Functional mixed effects models provide a robust framework for analyzing complex cytotoxicity data.
• This approach enhances the understanding of chemical-environment interactions.
• The methodology accelerates the process of setting environmental chemical regulations.