Table 5.

Reference-independent alignment accuracy, measured by TM-score, on the four benchmarks: In-House, MUSTER, SALIGN and ProSup

Methods	In-House	MUSTER	SALIGN	ProSup
HHpredL	1047.56	108.84	119.97	53.88
HHpredG	1522.77	142.00	121.83	56.44
MUSTER	–	136.47	–	–
BThreader	1537.89	143.95	132.85	66.77
CNFpred	1692.17	152.14	134.50	67.34

The result for MUSTER is its training accuracy (Wu and Zhang, 2008). All the other results are test accuracy. Bold indicates the best performance.