BAsE-Seq: a method for obtaining long viral haplotypes from short sequence reads

Table 1 Summary statistics from BAsE-Seq and Deep-Seq of hepatitis B virus

	Lib_1:9		Lib_1:99		S7.1
Read-pairs	14,352,128		17,083,497		42,997,995		8,197,770
Library	BAsE-Seq		BAsE-Seq		BAsE-Seq		Deep-Seq
Type of sample	Mixed clone		Mixed clone		Internal standard	Patient	Patient
Pass-filter read pairs^a	6,751,411 (47%)		8,816,934 (52%)		545,960 (1%)	26,066,408 (61%)	6,351,796 (77%)
Concordantly aligned^b	6,027,421 (89%)		8,150,721 (92%)		496,356 (91%)	23,366,358 (90%)	4,261,572 (67%)
High quality genomes	2,390^c		3,673^c		345^d	12,444^d
Type of analysis	Bulk	Individual	Bulk	Individual	Individual	Individual
Median per-base coverage depth	333,677	86	470,036	63	38	45	131,492
True SNVs detected	17 /17	17/17	15/17	17/17		68
SNVs detected							308
Errors detected	522	218	328	257	11
Highest per-base error	1.91%	0.202%	2.14%	0.231%	0.69%
Overall error	0.0524%	0.00674%	0.0324%	0.00541%	0.00214%

Each BAsE-Seq library was analyzed using the 'bulk' approach, i.e., without de-multiplexing by barcode identity, or the 'individual' approach, i.e., sequence reads associated with unique barcodes were analyzed separately. True SNVs in S7.1 were identified by BAsE-Seq as those that occurred at >0.69% frequency.
^aRead pairs after removal of adaptor and/or universal sequences. For BAsE-Seq libraries, this only includes read pairs that carry a barcode.
^bBoth reads in a pair were aligned in the expected orientation.
^c≥ 4 unique reads per base position across ≥85% of the genome.
^d≥ 4 unique reads per base position across ≥50% of the genome.

ISSN: 1474-760X