Table 4. Analysis of optimal alignment settings for each variable region when using full-length, region specific, and vertical-gap filtered full-length template sequences.

Region	Template sequences	Speed (seqs/s)	% Δsimilarity (sd)a	% Trimmedb
V19	Full-length	18	0.34 (0.64)	0.17
	Region-specific	NA	NA	NA
	Vertical-gap filtered	22	0.34 (0.65)	0.17
V14	Full-length	31	0.30 (0.84)	0.20
	Region-specific	37	0.31 (0.83)	0.20
	Vertical-gap filtered	41	0.29 (0.84)	0.20
V12	Full-length	51	0.29 (1.59)	0.29
	Region-specific	79	0.40 (1.52)	0.26
	Vertical-gap filtered	88	0.32 (1.57)	0.27
V2	Full-length	58	−0.09 (1.23)	0.02
	Region-specific	100	−0.01 (1.16)	0.10
	Vertical-gap filtered	105	−0.09 (1.23)	0.02
V23	Full-length	43	0.07 (0.91)	0.02
	Region-specific	64	0.14 (0.86)	0.23
	Vertical-gap filtered	65	0.07 (0.91)	0.02
V3	Full-length	69	−0.18 (1.50)	0.00
	Region-specific	122	−0.06 (1.35)	0.27
	Vertical-gap filtered	151	−0.16 (1.49)	0.00
V4	Full-length	61	−0.19 (1.00)	0.00
	Region-specific	100	−0.07 (0.75)	0.00
	Vertical-gap filtered	109	−0.19 (1.00)	0.00
V6	Full-length	78	−0.61 (3.63)	0.02
	Region-specific	145	−0.02 (2.92)	0.44
	Vertical-gap filtered	204	−0.64 (3.66)	0.02
V89	Full-length	45	0.09 (0.78)	0.13
	Region-specific	70	0.12 (0.75)	0.12
	Vertical-gap filtered	64	0.09 (0.78)	0.13
V9	Full-length	61	0.01 (1.24)	0.21
	Region-specific	100	0.08 (1.14)	0.17
	Vertical-gap filtered	102	0.01 (1.24)	0.21

^aSee description for Table 3.

^bThe percentage of sequences where less than 95% of the bases were aligned to the template sequence.