Table 1.

A spreadsheet representation of the investigation design graph shown in Figure 1.

Sample ID	Characteristics [Organism]	Characteristics [OrganismPart]	Protocol REF	Hybridization ID	ArrayDesign REF	ArrayData URI	Protocol REF	DerivedArrayData Matrix URI
liver 1	Homo sapiens	liver	P-XMPL-1	hyb 1	HG_U95A	Data1.cel	P-XMPL-2	FGDM.txt
kidney 1	Homo sapiens	kidney	P-XMPL-1	hyb 2	HG_U95A	Data2.cel	P-XMPL-2	FGDM.txt
brain 1	Homo sapiens	brain	P-XMPL-1	hyb 3	HG_U95A	Data3.cel	P-XMPL-2	FGDM.txt

Each initial sample has a Sample ID (the first column in the spreadsheet) and Characteristics – Organism (genus and species) and OrganismPart (the second and third columns). The terms used to annotate the characteristics can be obtained from the MGED Ontology [26], another suitable source of controlled vocabulary terms, or provided as user defined terms. The fourth column gives a reference to a relevant protocol, while the fifth gives the IDs of the three hybridizations performed. The reference to the array design type (HG_U95A) is given as a hybridization property, which is followed by the data file names, a reference to the data normalization protocol and the normalized data file.