Evaluating the usefulness of alignment filtering methods to reduce the impact of errors on evolutionary inferences

Table 1 Sensitivity and specificity of HmmCleaner

matrix name	c1	c2	c3	c4	global sensitivity	global specificity	sensitivity for 50 seqs	specificity for 50 seqs
default	−0.150	−0.080	0.150	0.450	94.26%	94.42%	98.42%	93.97%
species-rich	−0.175	− 0.175	0.150	0.400	85.70%	97.02%	97.36%	96.48%
high-specificity	−0.125	−0.125	0.175	0.400	88.89%	97.34%	96.64%	97.08%
species-rich and high- specificity	−0.125	− 0.125	0.150	0.400	74.03%	98.78%	94.11%	98.56%
HmmCleaner 1.8	−0.300	−0.100	0.200	0.500	94.56%	91.57%	98.74%	90.73%

The four new scoring matrices provided with HmmCleaner v2 and the scoring matrix equivalent to HmmCleaner v1.8 for comparison. c1-c4: values of the elemental scores for the four levels of residue conservation provided by HMMER. Global sensitivity and specificity were computed across all conditions of simulation, whereas the last two columns only used the most species-rich MSAs

ISSN: 2730-7182