Benchmark results for sequence comparison using the W-curve.

For each group, the number of sequences determines how many comparisons must be made, average number of bases in each sample can be used to estimate the processing rate in terms of the sequence lengths (vs. count of sequences). These times were taken from processing the upper-triangular distance matrices processed by neighbor and drawgram to produce the phenograms from the Mother/Infant study data [11]. Variations in rates are largely due to timesharing overhead, which reflects the likely environment of any clinical application. The data was measured using Perl-5.10.1 on linux-2.6.31 on an Asus M3N-HT (i.e., commodity desktop) motherboard with 4GB RAM and AMD Phenom 3.0 GHz processor.