10^th June, 2pm, Zeeman B3.03

24h physiological rhythms regulate most processes in the human body and their disruption is associated with many pathologies. The main challenge to studying such rhythms in humans is that they require multiple samples separated in time, i.e., a “time series”. This is infeasible from internal tissue that all house so-called circadian clocks. But are time series unavoidable? Here, I will describe a novel machine learning approach to reconstruct rhythms without time series. We apply this tool to public omics data to explore clock status in different human cancers.