Single gap penalty evaluation for a neutral (non-454 aware) model. The figure describes the alignment accuracy through Matthews Correlation Coefficient (y-axis) of the homopolymer insertion/deletion recovery versus the single gap penalty (x-axis). The accuracy is given for each of the seven degrees of difficulty, i.e. alignments against targets of an identity ranging from 100% down to 40% identity. Furthermore, the mean accuracy for all seven degrees of difficulty, at each gap penalty, is also shown by squares and a dashed line.
Lysholm BMC Bioinformatics 2012 13:230 doi:10.1186/1471-2105-13-230