Differential Gene Expression Feature Extraction from FHB Challenged Wheat RNA-seq Data.
Pan, Y., Li, Y., Liu, Z., Surendra, A., Wang, L., Foroud, N., Ouellet, T., Fobert, F. 2016. Differential Gene Expression Feature Extraction from FHB Challenged Wheat RNA-seq Data. Proceedings of the 8th Canadian Workshop on Fusarium Head Blight, Ottawa, ON, November 20-22 2016. P. 53.
In differential gene expression data analysis, one of our objectives is to identify groups of co-expressed genes from a large dataset in order to detect association between a group of co-expressed genes and a phenotypic trait. This has usually been done through various clustering approaches, such as k-means and bipartition hierarchical clustering based on certain similarity measures in the grouping process. In a differentially expressed gene dataset, the gene differential expression itself is an innate attribute that can be used in the feature extraction process. For example, a FHB differential expression gene dataset consist of one FHB susceptible line and n FHB resistant lines, each gene in each line would have three possible behaviors, up or down-regulated or unchanged after Fusarium challenge. We used three numerical values to denote such behavior, i.e. 1=up, 2=down, and 0=unchanged. As a result, we have up to 3n+1 differential expression patterns across all n+1 lines. Nevertheless, the actual number of patterns will be smaller than that since not all patterns would have a gene. This presentation is to demonstrate a series of successful applications of such feature extraction scheme in wheat gene differential expression data analysis under Canadian Wheat Alliance collaboration between NRC and AAFC.
Report a problem on this page
- Date modified: