A multiscale graph theoretical approach to gene regulation networks: a case study in atrial fibrillation.

The changes in between gene expression correlation structure induced in heart tissue by atrial fibrillation are studied by means of a graph theoretical approach. As expected by general statistical mechanics principles, the disease increases the general connectivity of the gene regulation network; the multiscale character of the analysis allows us to get both a general appreciation of regulation network connectivity and the sketching of a biological interpretation of the studied disease. The presence of a still largely unknown, scale invariant, global correlation field encompassing the entire genome is demonstrated as well.