[Hplusroadmap] A partial STM DNA sequencer and the $0 genomics project

[Bryan Bishop]

So, I took a look at DNA sequencing techniques, such as Sanger or the 
Maxam protocol, and I looked at how long it takes to sequence a few 
hundred thousand base pairs, and honestly I disliked what I saw. So I 
came up with my own method of doing DNA sequencing via scanning 
tunneling microscopes to measure the voltage difference between the 
probe tip (tungsten), the nucleotide, and some charged background 
surface. Now, typically STM machines are in UHVs (ultra-high vacuums), 
and if they aren't, then there accumulates this layer of water, dust 
and muck in general and it becomes unclean, which interferes if you 
want to look for a smooth surface. This means noise in the viewport, 
but if you're able to identify where your DNA is, then you're basically 
okay as long as you can continue to follow it -- maybe by assuming the 
molecule is completely straight (see molecular combing techniques).*

Notes are on the wiki at:
http://biohack.sourceforge.net/wiki/index.php/DNA_sequencer#Scanning_Tunneling_Microscopy_.28STM.29_DNA_sequencing

Other individuals have been able to construct STM machines for under 
$100, there's even a high schooler out there that was able to do 10 nm 
resolution (on a bumpy surface due to lack of vacuum, no less). So the 
technology is definitely doable and proven, and on the wiki I include a 
link to a ppt presentation from a group that was figuring out whether 
or not DNA sequencing is possible with STM, and their results say yes, 
but they didn't bother to try decoding a few bp -- this is peculiar. 
The work remaining for this DNA sequencer is basically assembling all 
of the resources on making an STM and repeating the results of the 
Washington group and beginning to write the software to use the piezo 
electric tip to follow the DNA molecule, plus the software to assemble 
the nucleotide information, and hey, maybe we can submit the 
first 'open source' sequence of data to GenBank or the other genome 
projects just for the hell of it.

Last week, Dan Bolser told me about the $0 genomics project existing out  
there somewhere on the internet. The origin of the project involves a 
*petition* to have governments start encoding human genomes and pay the 
costs of the sequencers. This is exactly what the roadmap calls for 
with respect to 'human archives' and so on-- storing the DNA of any 
individual, even the mind archives. A good start is storing community 
genomes. But, governments? We should take responsibility and do it on 
our own if (and indeed it is) necessary.

The $0 genomics project then, in my reinterpretation, is to build and 
promote the tools and equipment required to let communities set up 
their own RAID-farm and sequencer and to start collecting blood 
samples. Ideally, the system will be simple enough that librarians 
could set it up in a closet somewhere, cheap enough that it would apply 
to any university or any library or any community, and self-maintaining 
enough that it will be near zero maintenance. An interesting problem 
will develop with the STM-DNA sequencer, namely the different 
atmospheres where it would be employed, so perhaps a small vacuum 
chamber will have to be included with instructions on baking it, but 
this is a module problem and not something with the fundamental idea.

http://biohack.sourceforge.net/wiki/index.php/$0_genomics_project

Some ideas for the proliferation of this sort of project include asking 
the Internet Archive to host a genome bank, as well as the Long Now 
Foundation. Perhaps there are some other groups out there that would be 
interested in developing this technology.

Also, there are some alternative sequencing technologies that are 
showing up on my radar, such as the nanopores or artificial gels, but 
these require semiconductor manufacturing skills that, frankly, are 
incredibly hard to develop on your own. While I have enough information 
for basic semiconductor manufacturing on my website, it's basically 
millimeter technology at best.

- Bryan

* A note on eliminating noise. The background surface under the STM 
complicates things when you're in air. Namely, the surface interferes 
with whatever supersmall objects you have on the surface. A quick fix 
might be to have an anode and cathode hooked up on either end of the 
surface and allow DNA to crawl up and down the surface (before or after 
molecular combing). Then, use rapid scanning of the STM to detect 
motion, either by scanning while the DNA molecule is moving, or if that 
causes too much electrical interference, turn off the electrical 
current that is supposedly attracting the DNA molecule, and then update 
the known picture of the surface, which sounds like it will take a long 
time, but these surfaces are small and the scanning is quick. Once you 
find the DNA molecule, initiate the scanning sequence. From what I've 
read in STM operating manuals, HOPG surfaces can be bought cheap and 
wiped down with a special cleaning solution, and this is OK to expose 
to the air, and noise isn't too much of a problem without a vacuum.
________________________________________
Bryan Bishop
http://heybryan.org/