A clustering-based survival comparison procedure designed to study the Caenorhabditis elegans model.

Abstract

Caenorhabditis elegans is highly important in current research, serving as a pivotal model organism that has greatly advanced the understanding of fundamental biological processes such as development, cellular biology, and neurobiology, helping to promote major advances in various fields of science. In this context, the survival of a nematode under various conditions is commonly investigated via statistical survival analysis, which is typically based on hypothesis testing, providing valuable insights into the factors influencing its longevity and response to various environmental factors. The extensive reliance on hypothesis testing is acknowledged as a concern in the scientific analysis process, emphasizing the need for a comprehensive evaluation of alternative statistical approaches to ensure a rigorous and unbiased interpretation of research findings. In this work, we propose an alternative method to hypothesis testing for evaluating differences in nematode survival. Our approach relies on a clustering technique that takes into account the complete structure of survival curves, enabling a more comprehensive assessment of survival dynamics. The proposed methodology helps to identify complex effects on nematode survival and enables us to derive the probability that treatment induces a specific effect. To highlight the application and benefits of the proposed methodology, it is applied to two different datasets, one simple and one more complex.