Skip to main content
Fig. 6 | BMC Bioinformatics

Fig. 6

From: Constrained Fourier estimation of short-term time-series gene expression data reduces noise and improves clustering and gene regulatory network predictions

Fig. 6

Post-analysis of mouse T cell expression data. A The algorithm estimated the data with 2-harmonics Fourier approximation. The mean variance of the estimated frequency \(\omega\) between 3-replicates of each gene (log scale). 98% of the distribution had variance less than 0.01, indicating similar estimated frequencies between experiment replicates. B Selected TF activity predictions (using NCA) of noisy data (dashed) and Fourier de-noised data (solid). Replicates of Fourier estimated data are closely correlated (data on min and max cross correlation is given in Table 1). C Over 90% (29/32) of the TF activities had closer correlation (percent) with Fourier de-noised data than with noisy data. D Noisy data had exclusively higher mean angle between the replicates than the de-noised data, indicating that replicates of NCA predictions with de-noised data are more linearly dependent, and are closely related

Back to article page