Accuracy and quality of massively parallel DNA pyrosequencing

Table 1 Types of error

Error type	Number of occurrences	Percent of errors	Error rate
Insertions	58,337	36%	0.18%
Homopolymer extension and CAFIE	32,858	20%	0.10%
Not associated with homopolymers	25,479	16%	0.08%
Deletions	43,107	27%	0.13%
Incomplete homopolymer extension	13,868	9%	0.04%
Not associated with homopolymers	29,239	18%	0.09%
Mismatches	25,281	16%	0.08%
Homopolymer extension and CAFIE	16,725	10%	0.05%
Not associated with homopolymers	8,556	5%	0.03%
Ambiguous base calls (N)	34,184	21%	0.10%
Read errors	Number of occurrences	Cumulative percent of reads	Percent of reads
Reads with no errors (perfect match)	279,468	82%	82%
Reads with no more than one error	35,813	93%	11%
Reads with no more than two errors	11,651	96%	3%
Reads with more than two errors	13,218	100%	4%

Errors were classified as insertions, deletions, mismatches (substitutions) and ambiguous base calls (Ns). We further classified insertions, deletions and mismatches by their association with homopolymer effects. Deletions corresponding to an adjacent base are considered incomplete extension. Insertions and mismatches of the same base as an adjacent base are extensions. Insertions and mismatches that are the same as an upcoming homopolymer with no more than two intervening bases are considered carry forward errors.

ISSN: 1474-760X