Ultra-Cheap DNA Printing/Sequencing

[L]

Hey there,

We all would love to be able to print DNA as cheaply and easily as we can print a copy of a article. This would transcend printing just oglio primers. Every day I wish I could print whole genomes, and I'm sure most of you do as well. 

The issue: printing large DNA fragments costs about $0.35 a base, $0.10 at best. (Gen9)

This means that a 1 kb gene would cost $350 to make. Imagine the cost of printing 20 variants of that gene and testing the expression product of each to figure out exactly which version of your synthetic protein works best for you.  The cost would be insane. What if compiling a 350-line C++ program cost even just $50. Programming would be reserved to graduate schools, just as synthetic biology is today. 

Designing and successfully generating synthetic genomes in a garage for less than $500 per genome is necessary for the biotech revolution to play out even half as well as the computer revolution did. Thus, it's evident to all of us that DNA printing needs a massive cost reduction. 

While organizations like Cambrian Genomics and Gen9 are doing great work quickly, they are still mainly limited to academia even though I know that neither wants to be. (Especially Cambrian)

Perhaps it's time that we at DIYBio came up with our own means of DNA printing and sequencing. It may be rough and inaccurate at first, but I'm certain that the project will take off within a year or so, and eventually become industry-standard. It would be like the GNU/Linux of synthetic biology. ;) 

As a seed idea, we could use carbon nanotubes created using vapor deposition around nickle nanoparticles of around 12 nm in diameter. Then, in an argon atmosphere, the nanotubes would be doped with gold atoms on one side. DNA would be pulled through a stack of these nanotubes as in electrophoresis (from a positive charge to a negative one of the other side of the chamber) and surface interactions with the DNA bases could be monitored through the gold nanoparticles. It wouldn't be easy to pull off, but one divide could shotgun sequence DNA incredibly quickly with a high degree of accuracy. As for printing, DNA would be produced conventionally, but this time with a small laser at the end of the carbon nanotubes. If the DNA currently running through any given carbon nanotube did not match the desired strand, the laser at the end would activate, destroying the strand. 

Production of a device like that would not be something everyone was doing, but rather something in the public domain that any company could manufacture. We don't make our own CPUs, nor would they be easy to make at home (by not easy I mean virtually impossible), but we all know how to use them. Just like the modern CPU, however, we can do something that follows the same idea, just is much less efficient. From there, increasing purity would be all that is required. 

I would love to hear your ideas. 

-L

[Bryan Bishop]

On Fri, Feb 21, 2014 at 1:18 PM, L <lakopa....@gmail.com> wrote:

Perhaps it's time that we at DIYBio came up with our own means of DNA printing and sequencing. It may be rough and inaccurate at first, but I'm certain that the project will take off within a year or so, and eventually become industry-standard. It would be like the GNU/Linux of synthetic biology. ;) 

As a seed idea, we could use carbon nanotubes created using vapor deposition around nickle nanoparticles of around 12 nm in diameter. Then, in an argon atmosphere, the nanotubes would be doped with gold atoms on one side. DNA would be pulled through a stack of these nanotubes as in electrophoresis (from a positive charge to a negative one of the other side of the chamber) and surface interactions with the DNA bases could be monitored through the gold nanoparticles. It wouldn't be easy to pull off, but one divide could shotgun sequence DNA incredibly quickly with a high degree of accuracy. As for printing, DNA would be produced conventionally, but this time with a small laser at the end of the carbon nanotubes. If the DNA currently running through any given carbon nanotube did not match the desired strand, the laser at the end would activate, destroying the strand. 

I don't recommend using nanotubes on a first pass of an open source DNA synthesizer. Just build a typical machine- either with an array or the other type. The cost improvements can come later. The only open source DNA synthesis arrayer is approximately $20k in parts. That's not bad, but it's not great either. I am confident that a lower cost can be hit.

- Bryan
http://heybryan.org/
1 512 203 0507

[L]

Like I said, I was thinking that the synthesis reaction itself would occur traditionally using glass plated arrays. The resulting oglios would then be preferentially selected by sequencing all and destroying the ones that don't match the desired sequence.

So basically, yes. :)

[Cathal Garvey]

There have been a few papers on alternative reactions for DNA synthesis
that looked pretty promising, but I've never heard of them being
commercialised.

I wonder was that because someone tried them and failed to scale them
up, or because the risk involved is too high?

I can't believe it's the latter, because there's so much money to be
made with an alternative reaction that avoids the chain-branching issue,
but the former would surely have merited a follow-up paper..

Anyways, alternative reactions, if someone wants to try them, are where
the potential is I think. The phosphoramitide system is too prone to
side-branches, which limits the practical length of an oligo to only a
few hundred bases at most; I imagine most people make them only 100bp
long or so. Assembly and quality assurance are what add most of the cost
to the final gene; that's why a few companies offer such cheap 500bp
synth runs.

An alternative reaction without chain-branching, which can be recovered
as pure DNA using PCR, would be great. I've seen at least one paper on
this but lost it, and forgot the keywords! :( - It used a DNA analogue
for the synthesis step, which was compatible enough with DNA Polymerases
that it could be PCR'd into real DNA afterwards.

On 21/02/14 19:18, L wrote:
> Hey there,
>
> We all would love to be able to print DNA as cheaply and easily as we can
> print a copy of a article. This would transcend printing just oglio
> primers. Every day I wish I could print whole genomes, and I'm sure most of
> you do as well.
>
> The issue: printing large DNA fragments costs about $0.35 a base, $0.10 at
> best. (Gen9)
>
> This means that a 1 kb gene would cost $350 to make. Imagine the cost of
> printing 20 variants of that gene and testing the expression product of
> each to figure out exactly which version of your synthetic protein works
> best for you. The cost would be insane. What if compiling a 350-line C++

> program cost even just $50. *Programming would be reserved to graduate
> schools, just as synthetic biology is today. *

--
Please help support my crowdfunding campaign, IndieBB: Currently at
26.2% of funding goal, with 20 days left:
http://igg.me/at/yourfirstgmo/x/4252296
T: @onetruecathal, @IndieBBDNA
P: +3538763663185
W: http://indiebiotech.com

[Patrik D'haeseleer]

Always useful to read Rob Carlson's take on this issue:

Time for New DNA Synthesis and Sequencing Cost Curves

Patrik

[John Griessen]

On 02/21/2014 01:40 PM, L wrote:
> The resulting oglios would then be preferentially selected by sequencing all and destroying the ones that don't match the desired sequence.
>
> So basically, yes.

But, nanofabbing of nano-lasers, or even lenses that can aim at a nano speck of anything?
Not in my bag of tricks...

Cathal Garvey wrote:
> An alternative reaction without chain-branching, which can be recovered
> as pure DNA using PCR, would be great.

But, following a published method like,

"I've seen at least one paper on this but lost it, and forgot the keywords! - It used a DNA analogue

for the synthesis step, which was compatible enough with DNA Polymerases
that it could be PCR'd into real DNA afterwards."

seems less imaginary. Thermo-cyclers we can do. Nano-lasers NOT (yet...)

[Cathal Garvey]

Well, Sung (@bookhling) can do nano-lasers. I recall he did bio-nanodots
and got them to lase. Crazyballs.

[Nathan McCorkle]

you don't need some new kind of laser to do that though, you just need
the right optical know-how and a decent detector setup. Read the
patents by Helicos and you'll find plenty of similar setups in
academic literature:
https://groups.google.com/forum/#!topic/diybio/Ke7BqPUD46A

You can label-free detection by moving to a Raman setup.

--
-Nathan

[Cathal Garvey]

Hm, I happen to know a Raman expert! :)
For me though, while better chemical methods and nanotechnology stuff is
great, I'm a sucker for "self-bootstrapping systems". That's part of
what makes biology awesome, after all.

So, I won't be happy until we have a biological DNA synthesis reaction
or system that can be home-brewed in its entirity! Everything until then
is a stopgap!

26.4% of funding goal, with 20 days left:

[Nathan McCorkle]

With raman detection and sufficiently dilute dNTPS, you could
potentially use terminal transferase. Sourcing the dNTPs would be akin
to the recent thread on purifying the triphosphates... I remember you
comparing that to cleaning up agar into agarose at home, and how it
wouldn't really be efficient!

On Fri, Feb 21, 2014 at 5:10 PM, Cathal Garvey

--
-Nathan

[Cathal Garvey]

It's not important whether it's competitive with mass-market reagents,
only that it *can* be done.

For example, many programming languages can be "self-hosted" by
compiling their own compiler using a bootstrapping midway compiler. So,
bootstrap compiler is written in C, and is then used to compile the
language's own compiler from native source. Golang is like this.

But almost nobody does this; it's only important that it can be done if
necessary.

Likewise if the foundation of your revolution in synbio is utterly
dependent on one of a handful of chip-fabs around the world to make for
you, then I don't see it as complete. But once you've got a fallback
that you can always rely on if necessary, you continue to use the
cheapest, fastest, most efficient method.

TL;DR: I buy my agarose, but I'm glad I figured out how to make my own
if I'm ever stuck on an island with a desperate need for molecular
biology (oh wait, I am! One without ready access to reagents, then).

[L]

As for Ramen scattering and nanoscale lasers,

Nanoscale lasers can be accomplished through focusing two photons in 180 degree opposite phase at a single object. Destructive interference would cause the two photons to form an incredibly thin beam. 

I like the idea of Ramen spectroscopy, given that each base has its own highly unique vibrational signature. They probably do, but I personally don't know. 

The reason I recommended using carbon nanotubes is because they have a highly controllable size and can be manufactured (relatively) easily. It makes them a nice candidate for such a small chip. 

Finally, if this chip became a reality, I wouldn't say it would be limited in its scope to impact the world if it were only manufactured by a few specialized facilities in the world. Again, virtually no one can make their own CPUs yet the modern computer is essentially the foundation of our civilization at this point. This is because it totally shifts who can produce data. The Internet world isn't about producing your own hardware, it's about producing your own software. With biotech, it's about producing your own wetware, which is a very convenient way to do both. At this point, though, it costs far too much to do even the simplest things. Making wild-type E. Coli glow should cost $1 at max. This chip would allow for that. 

I'm not saying its the end-all, though. It's just the beginning. With a whole world of bioengineers, the cost of synthetic biology will drop exponentially as they build on each other's technology. Eventually, a device that can do everything I want the sequencer/synthesizer we're talking about here to do and so much more could be produced using only a small culture of a few novel cells. Our sequence/synth machinery will eventually be made by what our sequence/synth machinery originally created, just as robots make robots today. 
BUT, for that to happen, we need a world already highly involved in synthetic biology. Kind of like chicken and egg, no?
So that's why I say we figure this out. Because it will enable us to make the next generation of chips that can be made at home. It's a process we have to take one step at a time. What's really cool is that each step is exponential. 

-L

[Nathan McCorkle]

On Sun, Feb 23, 2014 at 12:01 PM, L <lakopa....@gmail.com> wrote:
> As for Ramen scattering and nanoscale lasers,
>
> Nanoscale lasers can be accomplished through focusing two photons in 180
> degree opposite phase at a single object. Destructive interference would
> cause the two photons to form an incredibly thin beam.
>
> I like the idea of Ramen spectroscopy, given that each base has its own
> highly unique vibrational signature. They probably do, but I personally
> don't know.

They do (old paper using UV excitation), sequencing has even been done
with Raman (Intel) based on concentration into and out of a PCR
reaction center.

> The reason I recommended using carbon nanotubes is because they have a
> highly controllable size and can be manufactured (relatively) easily. It
> makes them a nice candidate for such a small chip.

But you still have the issue of assembling them, and you haven't
really given a reason why a nanotube would be better than a nanopore
(which is much much more readily machined with something like a FIB,
or purchased off-the-shelf from SimPore, and integrated into a
layer-based microfluidic).

>
> Finally, if this chip became a reality, I wouldn't say it would be limited
> in its scope to impact the world if it were only manufactured by a few
> specialized facilities in the world. Again, virtually no one can make their

This sort of thing is getting quite common though, fabless IC
manufacturers are becoming more and more popular in everyday industry.

> own CPUs yet the modern computer is essentially the foundation of our
> civilization at this point. This is because it totally shifts who can
> produce data. The Internet world isn't about producing your own hardware,
> it's about producing your own software. With biotech, it's about producing
> your own wetware, which is a very convenient way to do both. At this point,
> though, it costs far too much to do even the simplest things. Making
> wild-type E. Coli glow should cost $1 at max. This chip would allow for
> that.
> I'm not saying its the end-all, though. It's just the beginning. With a
> whole world of bioengineers, the cost of synthetic biology will drop
> exponentially as they build on each other's technology. Eventually, a device
> that can do everything I want the sequencer/synthesizer we're talking about
> here to do and so much more could be produced using only a small culture of
> a few novel cells. Our sequence/synth machinery will eventually be made by
> what our sequence/synth machinery originally created, just as robots make
> robots today.
> BUT, for that to happen, we need a world already highly involved in
> synthetic biology. Kind of like chicken and egg, no?

I don't think so, the whole world wasn't drooling for computers and
yearning for some HP-type company to make it big in their garage...
the guy's starting HP in their garage were building the egg before a
chicken existed because they had a greater vision (whether it was to
further their own goals/desires or they were going for world-reaching
impact).

> So that's why I say we figure this out.

I'm less inclined to open-source something like this immediately,
since I've been incurring debt working on these and related ideas. I'd
have no problem doing that if I got something up and running, was able
to pay off my school debt and pool some money for further projects...
but at this point I'm quite wary of openly talking about the
secret-sauce I've been working on to this extent, as I genuinely fear
patent-trolls and companies that have the money to implement and take
to market stolen ideas.


I hope to have a nice relatively automated system up and running for
fabricating microfluidics sometime this summer, and I genuinely feel
that there is little need for new tech development, rather putting
existing pieces together in a novel fashion. It really is like how
Cathal was saying that no one has investigated agar purification
resin, or colicin self-selection... basically some people can only do
1 or 2 things in life, the people that might take longer but can do 3
or 4 things will be much more equipped to innovate. I.e. a biologist
can only count cells or pipette fluids, but combine that with computer
science and now they're a bioinformaticist or DNA gene function
hunter. If you add on Electrical Engineering or Optics, the scope and
possibilities for what they can accomplish grow tremendously (at the
cost of their increased time to learn, and often increased time being
a poor student, which arguably no one really enjoys).

There are lots of simply tricks buried in known science and
engineering, stuff that's so simple it blows your mind when you
realize it's beauty. Like all the crazy equations we used to have for
determining star future positions, all became much more unified with
some trickery of calculus. Sure the per-equation math got harder, but
the number of distinct equations dropped such that it more than offset
as far as cost-benefit.

[L]

Nathan, 

> But you still have the issue of assembling them, and you haven't

> really given a reason why a nanotube would be better than a nanopore.

Nanotubes are just one concept out of many. I have ideas for your concerns, and we'll get to those in a minute.

> The guy's starting HP in their garage were building the egg before a 

> chicken existed because they had a greater vision.

We have a greater vision here as well, just as the founders of HP did. While the whole world is drooling over GMO foods gone wrong, the field of biotech is pushing all kinds of boundaries. Plants as street lights, sutures made of DNA, et cetera. What the world wants is a horror story, like how they want robots to take over the world. Things like the Terminator Scenario sell well, as does the concept of Franken-food. Those of us who aren't so inclined to fear mongering, as our friends in the press seem to enjoy so much, don't have to worry about what the world is drooling over, just as all the great pioneers of computing didn't. 

> There are lots of simply tricks buried in known science and 

> engineering, stuff that's so simple it blows your mind when you 

> realize it's beauty.

I completely agree, and that is exactly the kind of technology we want to build this chip on. That's why I bring this up now after staying quiet for so long. I realized this idea of a read/write chip for DNA was getting no where within the University I'm at because of how small and perhaps over-academic the pool of people is, so I'm opening it up. Let people try to patent it; we'll slap an open-source license on it faster than they can say "patent pending." The only way for a community to get things done is through trust, and I trust all of you. 

So, back to the true issue at hand: The actual read and write mechanisms. Let's try building a small list of options first, and then go through each and pick a top two in each category. 

I'm partial to using carbon nanotubes as reaction centers because, as I said before, their size is easy to control, they are easy to make, and their properties are fairly well known at this point. We can use a static field to control their position on a chip down to nanometers, which is exactly what we need. The experimental setup wont be cheap, but we can wrangle something up and it's not like anything will be, that's the problem. A nanopore has no length, so it wouldn't be all too useful as the actual reaction center of either read or write, though I like the idea of a layered microfluidic system. The challenge there would be to have DNA held in known locations, but that's something we can surmount.

Raman spectroscopy and gold nanoparticles are both up for the read "head," but neither would be easy to do. Other ideas? 

-L

[Bryan Bishop]

On Sun, Feb 23, 2014 at 2:01 PM, L <lakopa....@gmail.com> wrote:

Again, virtually no one can make their own CPUs yet the modern computer is essentially the foundation of our civilization at this point

There are people who homebrew; it doesn't take overwhelmingly great skill.

[L]

Well good. That should make it clear that a digital read/write chip for DNA is plausible to create without financial ruin.

[Bryan Bishop]

On Mon, Feb 24, 2014 at 10:01 PM, L <lakopa....@gmail.com> wrote:

Well good. That should make it clear that a digital read/write chip for DNA is plausible to create without financial ruin.

But why bother making that as a first or initial implementation? As opposed to another implementation.

[L]

What do you mean? 

[Nathan McCorkle]

On Mon, Feb 24, 2014 at 7:54 PM, L <lakopa....@gmail.com> wrote:
>
> I completely agree, and that is exactly the kind of technology we want to
> build this chip on. That's why I bring this up now after staying quiet for
> so long. I realized this idea of a read/write chip for DNA was getting no
> where within the University I'm at because of how small and perhaps
> over-academic the pool of people is, so I'm opening it up. Let people try to
> patent it; we'll slap an open-source license on it faster than they can say
> "patent pending." The only way for a community to get things done is through
> trust, and I trust all of you.

I'm literally saying I want to do this to pay off my school loans, and
that I wouldn't want to release my plans, secrets, or sell devices.
I'd sell DNA as a service to begin, not at cost but likely
orders-of-magnitude cheaper than current market prices. I'm also
concerned with specific legal matters like ITAR (because I'm in the
U.S.) and, practical matters like nefarious powers whipping up
bioweapons.

The verbiage as-is is still a bit unclear when it comes to
nucleotides, as they are 'parts' of DNA.
"1. "Genetic elements"include, inter alia, chromosomes,
genomes, plasmids, transposons, and vectors, whether
genetically modified or unmodified, or chemically synthesized
in whole or in part."

Section XIV subsection (m) of ITAR doesn't look to promising
('Technical data') for DNA synthesizers, since polynucleotides are
mentioned in XIV subsection (g).
http://www.pmddtc.state.gov/regulations_laws/documents/consolidated_itar/ITAR_Part121.pdf

>
> So, back to the true issue at hand: The actual read and write mechanisms.
> Let's try building a small list of options first, and then go through each
> and pick a top two in each category.

Most of that is on wikipedia.

>
> I'm partial to using carbon nanotubes as reaction centers because, as I said
> before, their size is easy to control, they are easy to make, and their
> properties are fairly well known at this point. We can use a static field to
> control their position on a chip down to nanometers, which is exactly what
> we need. The experimental setup wont be cheap, but we can wrangle something
> up and it's not like anything will be, that's the problem. A nanopore has no
> length,

That's completely wrong, there is generally some depth associated with them.

> so it wouldn't be all too useful as the actual reaction center of

Sure it would be, the whole point is to know where your reactions are
happening and decrease diffusion times.

> either read or write, though I like the idea of a layered microfluidic
> system. The challenge there would be to have DNA held in known locations,
> but that's something we can surmount.

The whole point of down-scaling in the first place is to beat your
'challenge', stuff is smaller so your statistics of where it can be
decreases significantly.

> Raman spectroscopy and gold nanoparticles are both up for the read "head,"
> but neither would be easy to do. Other ideas?

Actually not really, you need a good interference notch filter though
($100-$300). http://www.youtube.com/watch?v=tRrOdKW06sk


--
-Nathan

[Nathan McCorkle]

On Mon, Feb 24, 2014 at 8:12 PM, L <lakopa....@gmail.com> wrote:
> What do you mean?
>

He means why not replicate macro-scale DNA synthesis first, but it's
orthogonal thinking and orthogonal tooling, and not worth it in my
opinion because you can do the same experiments on a meso or micro
scale just the same as the macro scale once your tooling is setup.

[John Griessen]

On 02/24/2014 10:24 PM, Nathan McCorkle wrote:
> I'm literally saying I want to do this to pay off my school loans, and
> that I wouldn't want to release my plans, secrets, or sell devices.
> I'd sell DNA as a service to begin, not at cost but likely
> orders-of-magnitude cheaper than current market prices. I'm also
> concerned with specific legal matters like ITAR (because I'm in the
> U.S.) and, practical matters like nefarious powers whipping up
> bioweapons.

So, you think open collaboration on synthesis technique using nano
elements to constrain DNA is doomed because of business and political climate?
Has to be in secret? Seems even more likely to be shut down and locked
in a closet along with the inventors if done in secret. To me at least.

There are practical difficulties with building anything with a nano part --
the tools available are complex and a high hurdle to consider using.
There are free tools for chip design though, that would let one specify
the micron scale microfluidic layers and the nanoscale zones where
a pore is wanted.

On 02/24/2014 10:01 PM, L wrote:> Well good. That should make it clear that a digital read/write chip for DNA is plausible to
create without financial ruin.

Well, maybe Bryan meant it's feasible to make one's own computer out of commodity parts,
but for nanoscale zones for novel reading/writing of DNA there're no prior building block
parts yet, are there? There *IS* going to be a high dollar hurdle to experiment in that
kind of machinery. You have some sensitive amplifiers and electrostatic power supplies
to connect to nano tiny elements, and to avoid noise ruining the effects of electronics,
the amplifiers and power supplies may need to be close to the nanoscale themselves, so
you are into chip fabbing costs. $40k per test run is what that costs.
When I said before, "nano is not in my bag of tricks", I meant practically, but I do have
experience there working for companies like Atmel, Cirrus Logic, Motorola, where the efforts
were channeled at mass market uses to be sure the development costs were recoverable.

So, to really crack this for open biotech uses still seems hard to me.

[Cathal Garvey]

Citation needed. If there's a practical way to make a *useful* CPU DIY
without "overwhelmingly great skill2, I haven't yet seen it. It would be
huge; a way to trust computers because you've built them yourself.

The last I saw of DIY-CMOS was an awful handful of gates that did
something less complex than a 555-timer. I don't see how that could have
progressed to CPU within a year.

--
Please help support my crowdfunding campaign, IndieBB: Currently at

27.7% of funding goal, with 17 days left:

[SC]

>I'm literally saying I want to do this to pay off my school loans

 

I have huge respect for anyone who starts a business to pay off school loans.  As a practical matter, I'm not sure this is the way to go. People that need synthetic DNA made for them typically don't have the resources to test it themselves, so you would have to do QA before shipping product.   More to the point, you would have to convince customers you could do QA properly and were reliable. Your end product is someone else's starting component for a more complex process, and if it's contaminated with incomplete strands (for example), your customers will waste time and resources and won't be happy.

 

As synthetic DNA isn't that expensive to buy from well established labs, I'm not sure that type of home business would be able to compete effectively in the current marketplace.   As a consumer of such things, I'd probably go with a better known supplier that came with recommendations from my colleagues rather than save a little money dealing with a home lab.   I don't mean to discourage you, so please just consider this a bit of marketing information from your target group.

 

However, if you're doing this for the spirit of invention and don't need to generate an income form it right away, I wish you the best of luck.

[John Griessen]

On 02/25/2014 10:06 AM, Cathal Garvey wrote:
> The last I saw of DIY-CMOS was an awful handful of gates that did
> something less complex than a 555-timer. I don't see how that could have
> progressed to CPU within a year.

Same here, which is why my first reply to this thread was, "Not in my bag of tricks."

An area of practical evolution in DIY computing, (and sensors also), is organic semiconductors
and room temperature processed inorganics to make fairly large flat circuits. Some
of the semiconductors may work OK on micron scales, but the DIY aspect shoots along
much faster, more effective when aimed at printed circuit scales, such as minimum
feature size of 100 microns.

Back to brainstorming on DNA reading or printing: However you get
electronics interfaced is going to mean going from macro scale with features
of 100 microns to nano scale with features of .1 micron. Some kind
of self assembling nano-pore-or-tube connected in sensor-useful ways
to flags of conducting material along with one nano-pore-or-tube might allow
everything else to follow. So, to my thinking, the problem is crossing scales
mostly, not conceptually how to detect a base pair once they are constrained tightly
so they don't wiggle away while measuring their presence.

What if you could have a bunch of sites that "attract DNA and have some electrodes attached"
located randomly on a plane? So that between them random lengths of DNA might attract to one or more?

Suppose you could sense if a strand were "attracted", and move it, but not "see" anything else?
Then, a way to sense if one strand were attracted" to two stations at once might be found...
by tugging on one end while holding tight to the other and seeing motion stop, or require
higher actuation volts for "tugging a strand" or ??

Anyone else thinking like this?

On 02/25/2014 10:14 AM, SC wrote:> As synthetic DNA isn't that expensive to buy from well established labs, I'm not sure that type

of home business would be able to
> compete effectively in the current marketplace.

Huh? Weren't we discussing how to reduce costs by 100X?

[Nathan McCorkle]

On Feb 25, 2014 8:14 AM, "SC" <stac...@yahoo.com> wrote:
>
> >I'm literally saying I want to do this to pay off my school loans
>  
> I have huge respect for anyone who starts a business to pay off school loans.  As a practical matter, I'm not sure this is the way to go. People that need synthetic DNA made for them typically don't have the resources to test it themselves, so you would have to do QA

The proof is in the pudding, though, if I sent Cathal/mega/Sebastian a vial of liquid... Any of them would quickly be able to if it coded for whatever they wanted. They'd be doing characterization on the DNA from their colonies anyway... PCR, restriction digests, and western blots are very common tools in any experiment trying to prove they cloned something or prove that something is expressed. Thesr are standard molbio tools, nothing new, nothing a decent cloning lab won't have. The only thing uncommon in diybio is the western blots, but that's why you just rely on a complex reporter system to act like a 'physical hash' of the DNA. If it smells like bananas and fluoresces, the DNA was good. Determining percent correct strands would be a bit more challenging, but again nothing a lab in any large city doesn't already perform everday (cytometry). Cytometry is pretty cheap sample wise, and most hospitals have a lab.

before shipping product.   More to the point, you would have to convince customers you could do QA properly and were reliable. Your end product is someone else's starting component for a more complex process, and if it's contaminated with incomplete strands (for example), your customers will waste time and resources and won't be happy.
>  
> As synthetic DNA isn't that expensive to buy from well established labs, I'm not sure that

Really? You're joking right? I'm talking about orders of magnitude cheaper than current markets, my initial goal is 100x cheaper, but my rough calculations for downscaling the reaction even further (and still keeping overall reaction yield low) show 10^12 cost reduction with a 1micron cube reaction center. <=$1000 genome

type of home business would be able to compete effectively in the current marketplace.   As a consumer of such things, I'd probably go with a better known supplier that came with recommendations from my colleagues rather than save a little money dealing with a home lab.   I don't mean to discourage you, so please just consider this a bit of marketing information from your target group.

No it's good to hear that when I get it right people's minds will be blown.

>  
> However, if you're doing this for the spirit of invention and don't need to generate an income form it right away, I wish you the best of luck.

Thanks, I was lucky to start studying on this my first year of college, and its been my goal since then. That was 2008, and I feel much more confident these days. Lots of engineering tricks have been accumulating in that time, lots of business networking, lots and lots of research and mostly instrumentation development.

Luckily I made friends with a guy who has a FIB and underused microfab capable wet lab... So things are almost in place for me to make so leaping progress in the coming months.

>
> --
> -- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
> Learn more at www.diybio.org
> ---
> You received this message because you are subscribed to the Google Groups "DIYbio" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
> To post to this group, send email to diy...@googlegroups.com.
> Visit this group at http://groups.google.com/group/diybio.
> To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/546ab571-ca11-4a6e-9642-5cd477633a81%40googlegroups.com.
>
> For more options, visit https://groups.google.com/groups/opt_out.

[Nathan McCorkle]

Oh and also I likely wouldn't start distributing raw DNA, it would likely be in a phage or episomal vector, maybe a BAC... Mailing stab cultures or lyophilized/dried culture.

[Cathal Garvey]

If you could ship DNA at 1/100 the costs I pay, I wouldn't care if it
had a 1/1000 error rate. That'd probably be my threshold for
correctional PCR, but at 1/100 the cost I could easily afford sequencing
and a few primers to amend errors.

Mind you, I'd still prefer if you'd do the QA at your end, and I only
pay 1/10 the price instead of 1/100, but someone could always operate a
middle-woman service to that effect and I could buy through them.

For me, DNA synth costs are a huge, huge factor in my ability to work,
so that'd be a huge change.

--
Please help support my crowdfunding campaign, IndieBB: Currently at

28.4% of funding goal, with 16 days left:

[John Griessen]

On 02/25/2014 03:27 PM, Nathan McCorkle wrote:
> Luckily I made friends with a guy who has a FIB and underused microfab capable wet lab... So things are almost in place for me to
> make so leaping progress in the coming months.

Wow! For the rest of you, FIB means focused ion beam, which lets you blast away interconnect metal
in a silicon chip surface if you have it exposed and in the FIB mill vacuum chamber. That lets you
tweak things at low costs compared to new masks for chip fabbing. It might also let you carve
some increased opening in a glass or silicon channel used for microfluidics, but probably
not much use for one that is under glass, (complete and covered), only for ones that are half made.

[Nathan McCorkle]

Yep, which is where layering or multi-step processing comes into play.
You could for example make a microfluidic then take it into the FIB to
add nano features. Unfortunately the materials for micro and nano are
often incompatible without more intermediary steps (such as making
negatives and positive reliefs or your 3D surface to change to a
compatible material, or metallization), but if the nano features are
simply a layer of a sandwich, you just have to have good enough
alignment between your sandwich layers. Alternatively there are more
powerful FIB sources (the ion source for milling) that also happen to
be local to me, which can mill from the micro scale all the way to the
nano, while most FIBs slow down tremendously on larger beam spot size
this one actually increases. It's still slow though, compared to
exposing with photolithography and thinking of a nano intermediary
layer, and bonding layers is something require of any microfluidic, so
adding a layer doesn't really require new skills as far as bonding
steps.

With a layered approach, you can use off the shelf nanopores (aka
aperatures, tem windows) or nanotubes if you can get them to easily
polarize and they're dilute enough to be discrete tubes or clustered
groups of tubes... and get away with microlithography using a
microscope and projector or blu-ray writer optical sled. I'd like to
get this out as a kit, with a spin-coater, it will be open-source.

>
> --
> -- You received this message because you are subscribed to the Google Groups
> DIYbio group. To post to this group, send email to diy...@googlegroups.com.
> To unsubscribe from this group, send email to
> diybio+un...@googlegroups.com. For more options, visit this group at
> https://groups.google.com/d/forum/diybio?hl=en
> Learn more at www.diybio.org
> --- You received this message because you are subscribed to the Google
> Groups "DIYbio" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to diybio+un...@googlegroups.com.
> To post to this group, send email to diy...@googlegroups.com.
> Visit this group at http://groups.google.com/group/diybio.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/diybio/530D1612.1030007%40industromatic.com.

>
> For more options, visit https://groups.google.com/groups/opt_out.

--
-Nathan

[Alex Murer [Open Biolab Graz Austria]]

Kind of off-topic, but does maybe anyone here know if the activation of the phosphoramidites has to happen right before using them, or can you prepare a solution which lasts for a certain (if so, what time?) amount of time?

[Nathan McCorkle]

My first guess is that you'd be limited by how dry you could keep the
solution... and also that the activator might open the base ring given
sufficient time (though I might be thinking of an subsequent step's
reagent).

> --
> -- You received this message because you are subscribed to the Google Groups
> DIYbio group. To post to this group, send email to diy...@googlegroups.com.
> To unsubscribe from this group, send email to
> diybio+un...@googlegroups.com. For more options, visit this group at
> https://groups.google.com/d/forum/diybio?hl=en
> Learn more at www.diybio.org
> ---
> You received this message because you are subscribed to the Google Groups
> "DIYbio" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to diybio+un...@googlegroups.com.
> To post to this group, send email to diy...@googlegroups.com.
> Visit this group at http://groups.google.com/group/diybio.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/diybio/b93ee03c-4192-47ee-8822-cce2512ec930%40googlegroups.com.

[Nathan McCorkle]

On Tue, Feb 25, 2014 at 3:33 PM, Nathan McCorkle <nmz...@gmail.com> wrote:
> My first guess is that you'd be limited by how dry

Sorry, dry meaning free of water, not liquid in general.

[L]

On Tuesday, February 25, 2014 2:00:26 PM UTC-8, Cathal Garvey wrote:

If you could ship DNA at 1/100 the costs I pay, I wouldn't care if it
had a 1/1000 error rate. That'd probably be my threshold for
correctional PCR, but at 1/100 the cost I could easily afford sequencing
and a few primers to amend errors.

That's essentially my thought process. It may be far cheaper for us to manufacture tons of DNA strands in parallel with comparatively low accuracy, and then have some kind of QC mechanism right before output. We need some kind of nanomaterial that can keep things relatively in place long enough to read a strand AND destroy it if it is not what we wanted. Nanotubes, nanopores, nano-whatever. Anything that can be placed somewhat easily. I still like nanotubes more just because they are easy to align in small quantities using static fields. It's when things start getting larger than a few microns that the problems start to develop.

For me, DNA synth costs are a huge, huge factor in my ability to work,
so that'd be a huge change.  

 That's exactly my point. All of a sudden, making many iterations of a novel genome you are kind of shaky on becomes easier than doing painstaking research for years to try to find a way to do everything you want to do using known sequences. The R&D time for each novel organism (I'm talking simplified bacteria here, not like an oak tree) would drop from a decade to months. Maybe even weeks for a well-funded lab. 
At this point, because of how cheap everything else would become as a result of having a DNA read/write chip, DNA synth costs are the limiting factor. 

[Nathan McCorkle]

On Tue, Feb 25, 2014 at 2:00 PM, Cathal Garvey
<cathal...@cathalgarvey.me> wrote:
> If you could ship DNA at 1/100 the costs I pay, I wouldn't care if it
> had a 1/1000 error rate. That'd probably be my threshold for
> correctional PCR, but at 1/100 the cost I could easily afford sequencing
> and a few primers to amend errors.
>
> Mind you, I'd still prefer if you'd do the QA at your end, and I only
> pay 1/10 the price instead of 1/100,

Would you do QC for me if I sent you the DNA for free? At the
price-point for actual product vs research costs, I should be able to
afford sending out free DNA to alpha testers.


Since the output of my system is a bio-packaged DNA, EU recipients
would be required to have a license technically, since it wouldn't be
naked DNA.

[L]

I would be down for that as well. Anything to get a cheaper synthesis system.

[Cathal Garvey]

Honestly, I'm not your guy for "partner" QC testing. I've done very
little sequencing in the past, and I'm not set up for QC levels of lab
cleanliness.

As an alpha tester, I'd eagerly receive free DNA with the caveat that
I'd have to offer feedback! :)
Especially as I usually order large CDS's, where the occasional
mutation, even an amino-modifying one, is not necessarily a dealbreaker.

[SC]

Curiously, have you priced out the materials?  Not the materials to make the device, but the materials to run it, the enzymes and nucleotides?

[L]

I haven't priced everything out yet, no. But dNTPs for 2500 50ul PCR reactions costs $210 from Qiagen. So given that and the few cents it costs to grow a yeast culture (to stitch DNA fragments together), I'm sure $300 per bacterial genome is fair, at least once the system is up and running.
Also, remember that once the chip works it would become trivial to produce your own enzymes for the reaction, just as it more or less is now with recombinant plasmids.

-L

[Nathan McCorkle]

Sorry I missed your message Stacy, yeah I've priced reagents from Glen Research, it comes to just under $1000, not including an argon tank, misc solvents, enzymes, cleanup kits, pipettors, scales, pumps, UV/vis spectrometer, etc... 

On Feb 28, 2014 5:15 AM, "SC" <stac...@yahoo.com> wrote:

Curiously, have you priced out the materials?  Not the materials to make the device, but the materials to run it, the enzymes and nucleotides?

--

-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/9ae988fa-a739-4a49-96ca-94374a7b9c40%40googlegroups.com.

[SC]

Cool! Best of luck, and let us know how it goes.