Database Reference
In-Depth Information
function}'', ''Biological process}'' and
''Cellular component}''). Each concept is
associated to its frequency in the cluster.
In this section, we have presented how XML
may be used to represent synthetic results of data
analyses. In the next section, we show how seman-
tic aspects that represent the domain knowledge
may be represented by metadata via ontologies.
In Figure 8, we can observe that cluster Clus-
ter_GSE6281_hclust_1 annotation contains the
two concepts GO:0005198 and GO:0043531
that are the most descriptive for it ; they refer to
'' cytoskeleton '' and '' nucleotide binding ''. As
stated above, the PMML representation keeps
the information on the number of occurrences
for each annotation and frequency in the cluster.
For instance 10 genes among the 26 genes into
the cluster are annotated by '' cytoskeleton '' and
4 only with '' nucleotide binding ''. Theses an-
notations provide the semantic meaning of each
cluster and are essential for a comprehensive of
the investigator.
Metadata on differentially expressed
genes and genes clusters
Ontologies provide an organizational framework
of concepts and a system of hierarchical and as-
sociative relationships for a domain. They are
useful for reuse and sharing and moreover the
formal structure coupled with hierarchies of
concepts and relations between concepts offers
the opportunity to draw complex inferences and
reasoning. In AMI, we reused the existing ontol-
ogy MeatOnto (Khelif et al., 2007) to annotate
Figure 4. Cluster_GSE6281_hclust_1
Search WWH ::




Custom Search