
Home : Workgroups : MAGE : Introduction : BioAssayData
BioAssayData
One of the central principles of MAGE is that data objects are regarded as 3dimensional matrices, where there are bioassays (experimental steps or conditions) along one dimension, design elements (spots) along the other dimension and quantitation types (e.g., signal intensity, background intensity) along the 3^{rd} dimension. Bioassay data objects can be represented in one of two ways: as a set of vectors in the form (value, dimension1, dimension2, dimension3) (useful for small amounts of data), or as a 3D matrix (BioDataCube). Transformations (e.g., filtering, normalization) can be applied to one or more bioassay data objects, resulting in derived data objects. A transformation involves computing values of the resulting 3D matrix from the values of source matrices, and it also transforms dimensions. On this diagram just the mapping of quantitation types into new quantitation types has been shown; DesignElementMapping and BioAssayMapping are modeled similarly. A quantitation type mapping transforms a list of quantitation types into another list of quantitation types, and it consists of maps that deal with single target quantitation types.
