Table 1.

Comparison of the performance of NER methods for sequence variants.

Method	Precision	Recall	F-score
OSIRISv1.1	0.97	0.3	0.46
OSIRISv1.2	0.99	0.82	0.89
Vtag	0.85	0.79	0.82
MuteXt*	0.87	0.57	0.67
MEMA	0.98	0.75	0.85

OSIRISv1.1 and OSIRISv1.2 were evaluated on the corpus described in this study. Performance of other methods correspond to published data for each development. Since the use of different benchmarks and the different evaluation criteria make the results not strictly comparable, the results can only be considered for a rough comparison of the state of the art. *Average of evaluation on two protein families reported