Skip to main content
. Author manuscript; available in PMC: 2016 Feb 3.
Published in final edited form as: Bioinformatics. 2007 Mar 25;23(11):1427–1428. doi: 10.1093/bioinformatics/btm095

Table 1.

Comparison of log-affine and affine gap costs

Lengths Log-affine Affine


Speed Scaled Quality Speed Scaled Quality
1600 × 1600 570 100 72% 211 37 58%
1600 × 800 285   50 78% 110 19 68%
800 × 800 147   26 69%   58 10 57%
800 × 400   76   13 76%   32   5.6 67%
400 × 400   41     7.1 64%   18   3.2 56%

Sets of 1000 simulated sequence pairs were aligned holistically using a match cost of 0, a mismatch cost of 1 and a gap cost of either wk = 1.67884 + 0.279091 k + 1.06161 ln k or wk = 6.86637 + 0.281073 k. Speed is the number of seconds Ngila spent aligning the sets, averaged across four runs. Quality is calculated by the average identity of Ngila’s alignments to the simulated alignments (Cartwright, 2006). Tests were done on a 3GHZ Xeon workstation.