A machine learning method to detect molecular species that oscillate in high-throughput circadian experiments. This method is based on BIO_CYCLE, a dataset including both synthetic and real-world biological time series, and both periodic and aperiodic signals. BIO_CYCLE estimates which signals are periodic in high-throughput circadian experiments, producing estimates of amplitudes, periods, phases, as well as several statistical significance measures. BIO_CYCLE is particulary useful to address several other related circadian problems, such as analyzing periodicity in high-throughput circadian proteomic data, or inferring sample time in different species.
Department of Computer Science, University of California-Irvine, Irvine, CA, USA; Department of Statistics, University of California-Irvine, Irvine, CA, USA; Department of Biological Chemistry, University of California-Irvine, Irvine, CA, USA
BIO_CYCLE funding source(s)
This work was supported by grants from the National Science Foundation (NSF IIS-1550705) and the National Institutes of Health (NIH DA 036984).