Highly improved homopolymer aware nucleotide-protein alignments with 454 data

BMC Bioinformatics

Table 3 Parameter evaluation using a 454-model using flowpeak information

Parameters/ID	40%	50%	60%	70%	80%	90%	100%
S = 30, D = 60, h = 0.2, k = 0.4, V = 1	0.681	0.682	0.708	0.748	0.775	0.796	0.815
S = 30, D = 60, h = 0.2, k = 0.5, V = 1	0.681	0.683	0.712	0.752	0.782	0.803	0.823
*S = 20, D = 40, h = 0.3, k = 0.4, V = 1*	0.673	0.677	0.706	0.757	0.788	0.812	0.833
S = 30, D = 60, h = 0.2, k = 0.8, V = 1	0.666	0.671	0.701	0.757	0.791	0.817	0.839
S = 15, D = 30, h = 0.4, k = 0.5, V = 1	0.613	0.629	0.673	0.744	0.782	0.825	0.849
S = 8, D = 15, h = 0.7, k = 0.4, V = 0	0.321	0.317	0.455	0.639	0.697	0.774	0.862

The table describes the Matthews Correlation Coefficient of the homopolymer insertion/deletion recovery for various parameter combinations using flowpeak information (i.e. SFF input). The results are evaluated using alignments against targets of a protein identity ranging from 40 to 100%, not including potential sequencing errors. The best results are highlighted in bold for each column as well as the parameter set which results in the best mean MCC (default parameters).

ISSN: 1471-2105