80 A Practical Guide to Cancer Systems Biology
Figure 4. The submission page of “profile clustering” analysis.many missing values will be excluded and the remaining missing values
will be imputed.- Specify the cluster number or select the automatic assignment. This is a
required parameter for fuzzy c-mean algorithm.^12 Users can set a specific
number or choose “automatic estimation” which will automatically assign
an optimized number by the iterative procedures (Fig. 4(b)). - Specify the parameters for function enrichment analysis. Select a back-
ground set and define ap-value cutoff (Fig. 4(c)). - Specify the parameters for motif enrichment analysis. This analysis
is performed by Motif-X algorithm.^13 Select a background set, set a
minimum number of a motif occurring in the foreground data, and define
ap-value cutoff (Fig. 4(d)). - Click on “Submit” to get to the result page.
- Inspect profile clustering analysis results. The resulting page reveals each
cluster by a profile plot (Fig. 5). Hit a specific profile plot to list the
members of the cluster as well as the overrepresented motifs and GO
terms. To visualize overrepresented GO terms, the GO terms and their
overlap can be organized into an enrichment map. For more details about
interpreting the enrichment map, please refer to the section “Construction
of function profile”.