Analysis Of Gene Expression Data Based On Wavelet Denoise And Cross-spectrum Evaluation

The Human Genome Project has led to a massive outpouring of genomic data along with the significant advances made by gene sequencing and DNA microarray techniques. Biology research has come into the era of information extraction and data analysis. It has been identified as one of the rapidly growing areas and central role of bioinformatics because of its vital role in genomics and proteomics. Processing of gene expression data has become the hotspot of bioinformatics.Background knowledge of gene expression data analysis was introduced in this paper. The current condition and existing methods of gene expression data analysis were elaborated. Relations between this topic and Electrical Engineering were discussed. Gene expression data were analyzed from signal processing perspectives.Solutions to inherent noise in gene expression data were introduced. A scheme was put forward based on wavelet denoise and fuzzy c-means clustering. Wavelet function selection and wavelet coefficients processing were discussed. Fuzzy c-means clustering method was introduced. Comparison clustering results of denoised and un-denoised gene expression data show that this method can effectively reduce background noise and enhance the accuracy of clustering.Then cross-spectrum evaluation method was applied to gene expression data in order to find time-delay relations among genes. The availability and feasibility of this method in theory was discussed and it has been used for gene expression data analysis. The simulation results indicated that this method can find time relations among genes and classify them correctly.Potential clues of some biological uncertain function genes may be provided from the above results. But much research still needs to be done in order to discover secrets of gene interactions. It is unpractical to construct total gene regulatory network because gene regulatory relations are far more complicated.