School of Public Health
In biomedical research a growing number of platforms and technologies are used to measure diverse but related information. The resulting data often constitute a collection of multi-way arrays in which one or more dimensions are shared. For example, on a common cohort of biological samples one may use genetic sequencing to identify single-nucleotide polymorphisms, and also use a microarray to measure the expression level of several genes at different time points; these data consist of a 2-way array (Samples X SNPs) and a 3-way array (Samples X Time X Genes) in which the dimension Samples is shared. Such datasets are often quite large, and methods that reduce dimensionality by exploiting common patterns are crucial for many applications. Several methods have recently been developed for the integration of multiple 2-way arrays (matrices). These researchers are extending this methodology to handle the integration of multi-way arrays in general.