The ultimate objective of this project is to exploit the emerging capabilities of deep sub-micron semiconductor technology to engineer ultra-fast, inexpensive, single molecule DNA sequencing. The basic concept was invented by Branton and Deamer in 1993 and first published in 1996. The initial implementation used a biological pore, alpha-hemolysin toxin from S. aureus in a membrane. Ions in solution were pulled through the pore of a protein channel by a voltage gradient. The ion flux (current) through the channel was measured using standard single-channel recording methods. Also in solution, and occasionally also traversing the pore but largely blocking the ionic current, were unfolded single-stranded RNA polymers. In a series of proof-of principle experiments, the strand lengths were shown to be proportional to the time the current was partially blocked. Long sequences of the same bases were shown to impede current by amounts differing with the largest and smallest bases. However, the noise and time response are such that individual bases cannot be detected with the present pore and electrode geometry. This proposal will study how capabilities just becoming possible in nanoscale semiconductor technology should make more optimized hole geometries possible. An additional advantage is that low noise detectors with greater than 10 MHz bandwidth and high gain can be placed right at the hole periphery maximizing charge and current sensitivity. The proposed device consists of a thin silicon membrane with a nanometer-sized hole. Charge passing through the hole is measured by a vertical transistor positioned along the wall of the hole via image charges in the transistor. The hole is made by standard techniques followed by narrowing via thermal or anodic oxidation. The vertical FET is built starting from a silicon-on-insulator wafer after several steps. The potential is for orders of magnitude improvements in speed, cost, and minimal pre-and post-measurement procedures compared to current sequencing techniques. The basic measurement process is at MHz rates. Even if degraded to a few 10s of thousands of bases per second, an individual's genome could be sequenced in days instead of years. If the study and initial prototyping are promising, the intent is to develop a commercial system as rapidly as funding allows.
| Bokhari, Shahid H; Sauer, Jon R (2005) A parallel graph decomposition algorithm for DNA sequencing with nanopores. Bioinformatics 21:889-96 |
| Bokhari, Shahid H; Glaser, Matthew A; Jordan, Harry F et al. (2002) Parallelizing a DNA simulation code for the Cray MTA-2. Proc IEEE Comput Soc Bioinform Conf 1:291-302 |
