Binary State Pattern Clustering

Binary State Pattern Clustering (BSPC) is a method for pattern discovery in gene expression microarray data that makes use of a perspective that interprets gene activity as being in two or more discrete functional states (e.g., on and off). In this method the gene microarray derived expression levels are first classified into putative states. The state data is then used in an unsupervised manner to identify biological sub-classifications and the genes responsible for their discrimination.

This website is a supplement to Beattie B and Robinson PN (2006) Binary State Pattern Clustering: A Digital Paradigm for Class and Biomarker Discovery, Journal of Computational Biology 13(5):1114-30). Here we offer additional details of the algorithm (including pseudo-code), some examples of its application to publicly available microarray data, and an implementation of the algorithm in C.

 

Contact: Brad Beattie and Peter N. Robinson