Predictive evo-devo: From theory to an empirical framework to model the origin of variation with implications for predicting evolution

There is overwhelming evidence that natural selection makes organisms adapt to their environments. However, we do not understand how organisms differ in their ability to generate variation on which natural selection can act on. In this project, we will address the origin of phenotypic variation by leveraging the rise of data-driven dynamic modeling. We will develop and implement an empirical framework that uses data-driven algorithms to learn relationships between past and future developmental timepoints from time-series data, uncovering the developmental rules underlying phenotypic variation. Our recent theoretical work suggests that these learned relationships can generalize across conditions, and we will test this using both simulated and available experimental datasets. By establishing an empirical approach to modeling the generation of new phenotypic variation, we aim to overcome a critical bottleneck in predicting future evolution - a central objective in evolutionary biology.