Discovery and assessment of gene-disease associations by integrated analysis of scientific literature and microarray data

The paper outlines a methodology and presents a tool to help biomedical researchers in interpreting complex experiments by automatically discovering gene networks and underlying biological processes (revealed by gene-expression patterns) that usually are extracted manually using existing tools. The proposed method, first, starts by mining specialized medical literature available on the Web to discover possible associations between genes and diseases. Discovered gene-disease associations are subsequently explored by analyzing abnormally expressed genes using microarray data analysis. Afterwards, relevant gene networks are built by clustering these genes on the basis of the similarity of their profile expressions in microarrays data. Finally, molecular, biological processes, cellular components and molecular functions, which may have a role in the disease, are pointed out by querying the Gene Ontology (GO) database. The methodology is illustrated by a case study on neuromuscular disorders.