Skip to main content
. 2009 Sep 17;10(9):R98. doi: 10.1186/gb-2009-10-9-r98

Figure 3.

Figure 3

Performance of GenomeMapper compared with that of other short-read alignment tools. (a) Runtime, measured as wall clock time between invocation and termination of the program, averaged from 10 independent tests with different random sets of 500,000 short reads from Est-1. The worst test was excluded from average calculations. Error bars indicate standard deviation. mm, gaps, and edit refer to the maximal number of mismatches, gaps and edit operations allowed. GenomeMapper was run with four different parameter settings: the serial version; the parallel version on four cores; the serial version merely aligning NIMS of length 13 or longer; and the parallel version aligning only NIMS of length 13 or longer. SOAP was found running on up to four CPUs instead of only one CPU, as configured with the command line (option -p). (b) Average sensitivity, measured as the percentage of aligned reads. Only GenomeMapper and SOAP can perform gapped alignments. (c) Average sensitivity of alignments, allowing three gaps and four mismatches with a combined maximum of four edit operations, measured as number of reads with gapped alignments. Fractions refer to the number of all reads with gapped alignments.