High-throughput sequencing: a failure mode analysis

BMC Genomics

Table 1 Failure mode categories Failed wells were distributed into each category based on observational data taken during sequencing pipeline procedures and manual evaluation of electropherogram traces.

Failure Mode	Trace characteristic	No. of sequencing reactions	Percent of all failed wells (Q₂₀ < 600)
Blocked capillary	Noisy or no data with a low signal intensity value (<100). Verified with capillary control results.	64	5.5
Low signal strength*	Noisy or no data with a low signal intensity value (<100) that is very close to or falls below the instruments detectable limit.	310	26.5
Mixed clone w/ vector sequence	Clean vector sequence followed by noisy data immediately after the cloning site.	137	11.7
Mixed clone, no vector sequence	Noisy data throughout the trace with sufficient signal intensity.	27	2.3
Low signal to noise ratio	Discernable sequence peaks with strong intensity background noise.	22	1.9
Excess Dye peaks	Large dye front usually followed by noisy data.	10	0.9
Hardstop	Abrupt end to good sequence.	2	0.2
Repetitive Sequence	Long stretch of repetitive DNA sequence that is followed by slippage in sequence or noisy data.	17	1.5
Homopolymer stretch	Long stretch of a single nucleotide followed by slippage in sequence or noisy data.	99	8.4
Poly A Tail	Stretch of Ts (template A) followed by slippage in sequence or noisy data.	484	41.3
Total		1172	100.0

* Preliminary failure mode, further broken down to final failure modes in Table 2.

ISSN: 1471-2164