METHOD AND SYSTEM FOR IDENTIFYING TIME-SERIES RELATIONSHIPS OF GENE EXPRESSION LEVEL USING SIGNAL PROCESSING METRICS

摘要

The present invention converts gene expression patterns measured in time-series experiments into discrete, time-invariant signals, which are then evaluated using digital signal processing techniques/metrics, such as discrete Fourier transforms, power spectral densities, coherence, transfer gain, and phase shift. The processing techniques of the invention take into account the time-series aspects of the data set and identify gene-gene associations without "a priori" limitations.