Skip to main content

Table 1 Processing benchmarks for GenTHREADER versus mGenTHREADER. The processing time required for the sequence-profile method GenTHREADER is compared to the profile-profile version of mGenTHREADER. The data was collected by running both methods on sequences from the Human proteome using 148 identical CPUs (1.3 GHz Pentium IIIs from the domain) over a period of 10 hours. The throughput was measured as the mean number of sequences with structural annotations per hour.

From: High throughput profile-profile based fold recognition for the entire human proteome

Mean sequences per hour on 148 identical CPUs 1236 230
Mean sequences per hour on 1 CPU 8.35 1.55
Mean time for 1 Human sequence on 1 CPU (min) 7.18 38.61
Estimated time for all Human sequences on 1 CPU (days) 159.70 858.24