| The research products in bioinformatic and biomedical fields and the resulting volumes of data and literature are extremely huge and growing in unprecedented rates. Moreover, the advent of Microarray technology has increased the volume of these resources tremendously. Gene expression profiles, generated from microarray technology, carry useful and demanded genetic information. This research explores new knowledge extraction methods by combining techniques based on gene expression profiles and textual data from the bioinformatics literature. Initially, a number of new and adapted gene selection techniques are proposed to extract very small subsets of informative genes from gene microarrys. The proposed methods are based on computing thresholds and discriminating capabilities of each gene. The next step applies text mining to biomedical texts to discover new relationships between informative genes and biomedical terms. This research focuses on discovering significant relationships between biomedical entities and selected informative genes. The experimental results are very encouraging and promising. |
