Applying text mining and machine learning techniques to gene clusters analysis

Genomic data clustering is receiving growing attention. However, finding the biological meaning of the clusters is still manual work, which becomes very difficult as the amount of data grows. In this paper, the authors present a few experiments applying text mining and machine learning techniques to help associate meaning to gene clusters. These experiments were applied to paper abstracts and interaction database data related to Saccaromyces cerevisiae genes both for identifying text content and for explaining the biological meaning of the gene clusters found. The results were compared to information published by experts in molecular biology and a number of relevant equivalences were found.