A Spurious Equilibria-free Learning Algorithm for the BlindSeparation of Non-zoer Skewness Signals
Neural Processing Letters
Biologically valid linear factor models of gene expression
Bioinformatics
Hi-index | 0.01 |
In this paper we apply three different independent component analysis (ICA) methods, including spatial ICA (sICA), temporal ICA (tICA), and spatiotemporal ICA (stICA), to gene expression time series data and compare their performance in clustering genes and in finding biologically meaningful modes. Up to now, only spatial ICA was applied to gene expression data analysis. However, in the case of yeast cell cycle-related gene expression time series data, our comparative study shows that tICA turns out to be more useful than sICA and stICA in the task of gene clustering and that stICA finds linear modes that best match cell cycles, among these three ICA methods. The underlying generative assumption on independence over temporal modes corresponding to biological process gives the better performance of tICA and stICA compared to sICA.