DNA sequencer
From Biohack
Contents |
Open source sequencer - the Polonator
This is basically the pyrosequencing method (Ronaghi). Ronaghi's research was based off of the behavior of DNA polymerase and the phospholuminescence based off of which nucleotide is added to a DNA template strand. Metal beads are coated with fragments of DNA molecules. These beads are added into wells on a giant microarray. Pyrosequencing works well with 200-400 bp per fragment. So you need a large number of wells, a visual method of recovering the data, and then processing all of the information. You have to get good coverage and overlap of many of the fragments, otherwise you wouldn't know how to put them together once you have all of the different fragments. This generates tons of information, apparently over a zetabyte, and even Google only expands by 20 petabytes/day (and $1 million per MapReduce in terms of electricity costs).
DNA sequencing techniques
Traditional methods of DNA sequencing (basically) suck.
Sanger method / Dideoxy method
- Sanger sequencing (1999)
- Sanger-Coulson Sequencing
- What exactly is Sanger sequencing?
- Sanger's dideoxy method [pdf]
- The Sanger success story
- Dry-Lab Demonstration of Sanger Di-deoxy Sequencing of Cloned DNA
- Ref: Sanger, F. , Nicklen, S. & Coulson, A.R. (1977).
DNA sequencing with chain-terminating inhibitors.
Proceedings of the National Academy of Sciences, USA, 74, 5463-7.
Maxam-Gilbert sequencing
- Ref: Maxam, A. & Gilbert, W. (1977).
A new method of sequencing DNA.
Proceedings of the National Academy of Sciences, USA, 74, 560-4.
Chain-termination methods
BAC-to-BAC / Map-based method
Whole-genome shotgun sequencing
Pyrosequencing
Nanopores
A nanopore is an opening in a wall through which DNA can pass, but only one nucleotide at a time. The pore can contain something such as an ohm meter or a nanolaser to detect what is being passed through it, or a way to add energy to separate the nucleotides. The main disadvantage of this method is that you need to be able to build 100 nm scale objects, and who has this sort of equipment laying around at home?
- There's also the scanning tunneling microscopy technique which relies on quantum tunneling of electrons that pass through the nucleotide within the nanopore.
- See nanopore sequencing @ Wikipedia.
Ohm meter
- Determine the amount of resistence to an electrical current presented by whatever is currently blocking (connecting) the nanopore together.
Artificial gels on silicon
- Semiconductor manufacturing -> build an artificial gel made up of silicon, allow the DNA molecules to run through this Si-Gel, and see what happens. This is basically using an artificial gel as a component of one of the earlier methods of DNA sequencing.
Nanolasers
Theoretical DNA sequencing techniques (brainstorming)
Bryan's first method
- See chat logs in #biology and with Blake Reiger and Kyle M. Loh (RR|Neuroscience in ##neuroscience) for explanation. Turns out that this method will require at least 57 years to sequence 3 billion base pairs (such as the human genome). A first version of this method required 4^(3 billion) strands of DNA, but a nifty revision was introduced such that an incorrect prediction does not lead to the inability to use that strand again. If there is a way to speed up ligation beyond so-called 5 minute Quick Ligation to perhaps a few seconds then sequencing can be reduced to a month -- to speed up ligation, we would have to be able to target DNA and launch enzymes directly at it with 100% accuracy.
Copy-copy-copy method
For each bp, add 1k bp of the same nucleotide. Run this by a laser, and record how long it takes until a flip is made. You would have to prime the DNA molecule with your own known oligo sequence so that you can judge by how fast the molecule is running past the laser, so that you can figure out where a few thousand A was originally three adenosines and not just one (this is one of the disadvantages of the technique, yes, but priming should be able to help). This is assuming that you can run DNA past the laser at a very specific location.
2D grid amplification
For each nucleotide, add a few thousand orthogonal nucleotides of the same type. Ideally, these orthogonal strands will make a 2D manifold which a laser could scan. For bonus points, use this in combination with the copy-copy-copy method so that you have 1k nucleotides across and 1k nucleotides down for each base pair. A noticably larger target to scan.
Wishful thinking
It would be exceedingly helpful to be able to amplify the size of each nucleotide. In other words, to make a DNA strand that is tens of thousands of times larger. Another helpful technique would be where we can (1) be sure that there is only one DNA replication enzyme acting and (2) where we can track the use of nucleotides from 'free space' to the new DNA molecule.
Scanning Tunneling Microscopy (STM) DNA sequencing
- Build an STM machine (see Bryan's notes on DIY STM (cost is less than $100 USD)). See STM for information on building a scanning tunneling microscope.
- Method of isolating, mounting and finding DNA on a surface suitable for STM (see Babak, John Lund, etc. below).
- Molecular combing
- Molecular combing page (vids)
- Molecular combing protocol
- Chromosome combing over at Wikipedia.
- IETS
- ...
- Molecular combing
- Can an STM use tunneling and voltage detection to determine the nucleotide?
- Note that some sources say that STM-DNA sequencing is "slow" but other sources say that an STM probe tip can scan over an entire genome in 10 seconds. Who should we believe?
- Direct Electronic Identification of Oligonucleotides with Inelastic Electron Tunneling Spectroscopy [ppt]
- jlund@ee.washington.edu, babak@ee.washington.edu,
- Center for Excellence in Genomic Sequencing
- Declan Ryan, Maryam Rahimi, John Lund, Ranjana Mehta, Babak A. Parviz, "Toward nanoscale genome sequencing"
- ...
- DNA is 2.6 nm wide and a nucleotide is 0.33 nm long. A high school student has been able to nearly see resolutions down to 10 nm, so perhaps this would be sufficient for voltage testing of DNA.
