The structure of MIAME

The MIAME structure is divided into two major sections:

• array design description;

• experiment description.

This reflects the fact that arrays are often manufactured independently of concrete experiments in which they are used. Therefore array designs can be described in a separate section, which is independent of the experimental part, and can then simply be referenced from the latter. In general, all parts of a MIAME description should be given unique identifiers in order to reference them in the corresponding sections of the experiment description.

While some elements of MIAME, especially protocols, are described as free text, most sections are required to use controlled vocabularies or external ontologies. This is essential in order to allow database queries and automated analysis on the data. Here, the MGED group has developed the ontology MO, which is recommended to be used whenever appropriate. If MO or other external ontologies are not feasible for the data to be described, MIAME requests the user to apply their own qualifiers and values in the format of:

(qualifier, value, source)

triplets, where qualifier specifies the name of the element, value its current instantiation and source the reference, where the controlled vocabulary is described. For instance,

(qualifier: 'organism', value: 'homo sapiens', source: 'NCBI taxonomy').

In the following sections the elements of MIAME are outlined in detail giving a 'checklist' for the user to verify the completeness of a microarray experiment description according to the minimum information required. More information about MIAME and software that is related to it can be found at

0 0

Post a comment