Interpreting leukemia subboundaries

The table of Subboundary Analysis output is shown automatically once you generate the analysis. If you closed it and wish to reopen the table, go to the Projects menu, choose table and select the "local leukemia boundaries: subboundary analysis results" table. Alternatively, right click on that analysis in the Results tab of the project window.

The table includes values and p-values for each of the subgraph statistics. The histograms show the distribution of each statistic, with the original statistic for the boundaries shown as a slim red bar.

Four of seven of the measures of "boundary-ness" were significant for the leukemia dataset. In essence, there were longer boundaries (higher length, fewer singletons, long path length between pairs of boundary elements) than expected by chance.

Statistic

Meaning

Value

interpretation

NS

number of subboundaries

significantly low

boundary-like

N1

number of singleton Boundary Elements

significantly low

boundary-like

Lmax

maximum subboundary length (number of linked BEs)

significantly high

boundary-like

Lmean

mean subboundary length

not significant

 

Dmax

maximum subboundary diameter

not significant

 

Dmean

mean subboundary diameter

significantly high

boundary-like

D/L

mean diameter-to-length ratio (indicates branchiness)

not significant

 

Subboundary diameter is the shortest path length between each pair of BEs in a subboundary.

So, there are significant areas of local change in leukemia. The next question is: are they associated with TCE injection wells?

Next Question