Vincenzo Vitelli

This paper introduces Rotational Component Analysis (RCA) to detect low-dimensional rotational dynamics in neural data and shows these dynamics constrain cue-evoked behavior and learning across species.

This paper develops a data-validated theory of natural selection after catastrophe, showing mean fitness relaxes inversely with time and follows Levenberg-Marquardt optimization in E. coli.

The paper introduces an algorithm to identify condition-dependent transcription factor binding sites by compressing promoter sequences into hyperletters, validated on E. coli data.

This paper introduces a neural-network based method to identify functional groups in microbiomes, showing it can distill complex communities into few groups that explain community function across different environments.

This paper develops an information-theoretic framework for model reduction in physics and biology, identifying relevant variables that predict future dynamics and demonstrating its application to atmospheric flows and cyanobacteria colonies.

The paper uses machine learning to model BMP-driven cytoskeletal dynamics in Drosophila development and shows its conservation in human neuroectoderm morphogenesis.