Biclustering algorithms have matured from their initial applications in bioinformatics, evolving towards different approaches and bicluster definitions, which makes sometimes hard for the analyst to determine which one of the available algorithms best fits her problem. As a way of benchmarking these algorithms, several quality measures have been proposed in literature. Such measures cover numerical aspects related to the accuracy, the recovery power or the capability of retrieving previous biomedical knowledge. However, biclustering apparently remains as an uncommon option for biomedicine analysis.
Here we review the impact of biclustering algorithms in biomedicine and bioinformatics with the object of measuring and understanding non-numerical aspects of biclustering algorithms focusing on citation-based statistics that can be relevant for their application on the domain. In order to achieve this, we performed analyses of the citations impact of several clustering and biclustering algorithms, and propose a methodology that can cover this aspect of biclustering usage.