Third Generation DNA Sequencing with a Nanopore - Nanopores: Sensing and Fundamental Biological Interactions - page 308

While trapping a DNA molecule this way may facilitate the high fidelity reads
required for sequencing it, the electrical noise associated with the ionic pore current
introduces ambiguity that adversely affects base-calling. If dielectric noise asso-
ciated with the membrane predominates for f > 4 kT DC m pDf 2

Thus, the membrane capacitance, C m , should be minimized to promote better SNR.
But even if the dielectric noise contribution is minimized, it is likely that thermal
and/or 1/ f noise will remain problematic. Thus, it is likely that base-calling will be
accomplished in a noisy environment, extracting the current associated with a
sequence of bases from a signal buried in the noise, which will ultimately affect
throughput of signal fidelity.
To make this more concrete, we can estimate noise and the affect on the
throughput. The data shown above indicates that the relative change in current
associated with

1 kHz, then I rms ¼

l -DNA blockading a pore is typically DI/I < 0.8, which translates

to a DI ~2 - 3 nA for a 2.2 nm diameter pore in 1 M KCl at 0.5 V, near the stretching

threshold. Therefore, to detect a molecule with SNR > 2, we need peak-to-peak

noise < 1.5 nA or an rms value of DI rms ~ 1.5 nA/8 ¼ 190 pA. For a bandwidth of

Df < 100 kHz, we estimate that DC m ~ 70 pF is required to detect a current signature

[ 44 ], which is easy to do. On the other hand, according to MD simulations [ 65 ], to
detect the difference between a C-G and A-T base-pair in a pore like this, we must
resolve a difference signal of DI rms ~ 2 pA. It is realistic

to suppose that the dielectric noise can be forced to satisfy this specification by
minimizing the capacitance [ 44 ] even for a bandwidth of 16MHz. However, if thermal
noisepredominatesthen I rms ¼

DI~ 20 pA or smaller so that

p

4 kT Df =R

2 pA . Thus, the rms-current noise
specification forces a bandwidth no greater than ~40 kHz, which limits the through-
put. From these extrapolations, we conclude that a single, solid-state nanopore in a
membrane engineered with state-of-the-art fabrication techniques would have
adequate frequency and noise performance for sequencing DNA, but the noise is
likely to compromise the throughput and base-calling fidelity.

Acknowledgments We gratefully acknowledge numerous contributions and our close collabora-
tion with Jiunn Heng, Chuen Ho and Greg Sigalov. This work was funded by grants from National
Institutes of Health [R01 HG003713A, PHS 5 P41-RR05969], the Large Resource Allocation
Committee [MCA05S028], the Petroleum Research Fund (48352-G6), and the National Science
Foundation [TH 2008-01040 ANTC, PHY-0822613 and DMR-0955959].

References

1. Lander ES, Linton LM, Birren B, et al (2001) Initial sequencing and analysis of the human
genome. Nature, 409(6822), 860-921.
2. Venter JC, Adams MD, Myers E, Li PW, Mural RJ, et al (2001) The sequence of human
genome. Science 291, 1304-1351.
3. Mardis ER (2008) The impact of next-generation sequencing technology on genetics,” Trends
Genetics 24(3), 133-141.
4. Metzker ML (2010) Sequencing technologies—the next generation. Nature Rev. Genetics
11, 31-46.