The content test was done on all four architectures, yielding similar
results on each architecture. One sequence (13955) took significantly
more time than the others in the BLASTP test, and another sequence
(14133) took less time than the others in the BLASTX test. However,
for BLASTN, the run time was fairly consistent for all five sequences.
The Cray CS6400 search times are shown in Table 
, with
the neighborhood word list sizes and number of database word hits for
each sequence.
Table: Content test on Cray CS6400
The results of the sequence length test for the SGI Challenge are
plotted in Figure . The results for other architectures
are similar. The confidence intervals are plotted in the figure.
(a) BLASTN
(b) BLASTP
(c) BLASTX
Figure: Length test on SGI Challenge: Doubling the length of
the sequence doubles the run time.
The views and opinions expressed in this page are strictly those of the page author.
The contents of this page have not been reviewed or approved by the University of Minnesota.