Multiblock Orthogonal Component Analysis - A Novel Tool for Data Integration

Multiblock Orthogonal Component Analysis (MOCA) is a new data analytics tool! MOCA is designed to be fast and transparent when analyzing multiple .. Read more blocks of data registered for the same basis set of observations. It is similar in scope to O2PLS in cases involving only two matrices, but generalizes to situations involving more than two matrices without giving preference to any particular block of data. MOCA will extract two sets of components; joint and unique components. More specifically, the components may express globally joint information, locally joint information, and unique information: - Globally joint information is systematic structure found in all data blocks being analyzed; - Locally joint information is systematic structure found in a subset of the data blocks; and - Unique information is additional systematic structure found only in one, single data block.