DLP projector based DNA synthesizer proposal

[Gene Hacker]

Sorry if this is a repost, I just noticed the formatting was horrible
and hard to read on my last post.

Let's face it, materials for diybio are expensive. Enzymes and genes
cost A LOT. We need to be able to make our own. So what if we could
simply look up enzymes and/or genes we need online and download them
to real life? What if we could write genetic code on the computer and
have it compiled in real life? Well with a device called a DNA
synthesizer we can. Witnessing the success of the keiki gels, I
propose we have a project to make a DNA synthesizer. I propose we make
a maskless lithography based DNA synthesizer. Maskless
photolithography based gene synthesizers are quite fast in comparison
to other synthesizer and also enable us to make DNA microarrays.

DNA synthesis by maskless lithography works as follows: Put down a
layer of photo capping chemical. Use a digital micromirror chip to
shine light to remove the capping chemical from selected locations.
Put down one type of nucleotide, nucleotides bond to previously put
down nucleotides where the capping chemical isn't. The photocapping
chemical binds to DNA strand and prevents oligonucleotides from being
added to the strand. Lather, rinse, repeat and you've got yourself a
DNA chip with thousands of different sequences on it.

With this sort of DNA synthesizer we could definitely make DNA
microarrays, chips with a bunch of DNA strands on them used for
testing for certain sequences of DNA are present. IE, you can use it
to identify different organisms or figure out which variant of the flu
virus you have. It is also possible to extract the DNA strands from
DNA microarrays and combine them into longer DNA strands using the
process described here: http://arep.med.harvard.edu/pdf/Tian04.pdf

Doing maskless photolithography DNA synthesis requires a special chip
called a digital micromirror device(DMD), lucky for us these chips
come handily packaged along with some optics and high quality mercury
vapor lamp in DLP projectors. It is also possible to do maskless
photolithography with a DLP projector hooked up to the camera port of
a trinocular microscope: http://www.physics.rutgers.edu/ugrad/387/388s06/film_deposition/Musgr...
Using this method the group was able to achieve a smaller than the
necessary feature size for maskless photolithography DNA synthesis(<20
microns). Though they had trouble achieving uniform illumination which
means they had to adjust the brightness with an image overlay and
ended up only being able to use only 95% of the image after the
overlay was applied. However uniform brightness may not be as
important for DNA synthesis as it for semiconductor photolithography.

A typical resolution of a DLP projector is about 768X1024, this gives
us 786,432 pixels, with 80 base pairs(bp) per pixel and an overlap of
40 bp we can synthesize a DNA strand about 31,457,280 bp long with
reasonable accuracy. This is quite a lot. In comparison, the genome
of Mycoplasma genitalium(the bacteria that Craig Venter is modifying
for synthetic biology) is 580,073 bp long(source:
http://users.rcn.com/jkimball.ma.ultranet/BiologyPages/G/GenomeSizes....).
In other words we could synthesize the Mycoplasma genitalium genome
over 54 times over. 31,457,280 bp is a lot longer than any enzymes
we'd ever want to synthesize, in fact I don't think there are any
genes that are larger than 31,457,280 bp, heck I don't think there are
any bacteriophages with genomes larger than 31,457,280 bp.

However, the real trouble in making a DNA synthesizer is not the
synthesizer but obtaining the chemicals necessary, the price of
chemicals necessary(specifically the phosphoramidite nucleosides) is
ask for quote, which translates to very expensive.