Cambrian Genomics

[Koeng]

Hey guys I just wanted to know if anyone knows how far along this has gotten. I saw a recent article 

http://www.extremetech.com/extreme/153087-dna-laser-printing-heralds-new-day-for-genomics-research

But I really haven't seen anything new and recent :/ Seems like a cool project!

[Sumon Sadhu]

They just closed several million dollars in funding from patient and forward thinking silicon valley investors. 

It takes time to build something great so let anselm, Austen and co do their work. 

--
-- 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?hl=en.
To view this discussion on the web visit https://groups.google.com/d/msg/diybio/-/NCFQ12T_8SwJ.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
Sumon Sadhu

Chairman & Co-founder,
Imperial Entrepreneurs
t: +44 7854 153 996
e: sumon...@imperial.ac.uk

www.sharpshoot.blogspot.com
Skype ID: sks2sly4u2c

[Avery louie]

Can't wait.

--A

[Koeng]

Me either !!!!!

If I ever get that system, I am going to print every dna modification enzyme there is, then start creating synthetic genomes ":D

On Wednesday, April 24, 2013 8:37:54 PM UTC-7, Avery wrote:

Can't wait.

--A

On Wed, Apr 24, 2013 at 11:37 PM, Sumon Sadhu <sumon...@gmail.com> wrote:

They just closed several million dollars in funding from patient and forward thinking silicon valley investors. 

It takes time to build something great so let anselm, Austen and co do their work. 

On Wednesday, April 24, 2013, Koeng wrote:

Hey guys I just wanted to know if anyone knows how far along this has gotten. I saw a recent article 

http://www.extremetech.com/extreme/153087-dna-laser-printing-heralds-new-day-for-genomics-research

But I really haven't seen anything new and recent :/ Seems like a cool project!

--

-- 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+unsubscribe@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+unsubscribe@googlegroups.com.

To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio?hl=en.
To view this discussion on the web visit https://groups.google.com/d/msg/diybio/-/NCFQ12T_8SwJ.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Bryan Bishop]

On Wed, Apr 24, 2013 at 10:37 PM, Sumon Sadhu <sumon...@gmail.com> wrote:
> They just closed several million dollars in funding from patient and forward
> thinking silicon valley investors.

I had an informant living with Cambrian Genomics and I never got this
tiny detail:

"Cambrian Genomics brings in lasers only once the plate is covered
with many thousands of DNA-carrying beads, and once each of the beads
has been sequenced. With so many copies made, some predictable portion
will have been made error-free, and an automated laser flits about
over the plate and blasts any beads with a desired sequence off of the
plate and into a collector."

Using a laser to avoid contamination was the missing detail...
otherwise Cambrian's idea doesn't seem interesting.

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

[Nathan McCorkle]

Yep, see here for laser specifics:

http://www.youtube.com/watch?feature=player_detailpage&v=7yscphwaWNs#t=707s

--
-- 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?hl=en.

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

--
-Nathan

[Anselm Levskaya]

Hello all, 

Cambrian is very real.  We're making lots of progress, fast, but doing hardware/bio work is definitely a slower game than software development!

@bryan : this is not an 'interesting' approach.  It's a practical one informed by decades of experience building synthetic DNA constructs and optical systems. i.e. one that might actually work. ;)

Brief synopsis:

The basic problem of synthesis is that serial programmatic polymer synthesis is intrinsically error-prone.  Per-cycle monomer coupling efficiencies are never above 99.5% and generally more like 99% for microarrays.  This means that for a 100mer the percentage of correct, full length oligo molecules is only (0.99)^100 = 0.36 -> 36%  Two thirds of the stuff made at this length has deletions.  If you assemble this pool your synthetic gene will be riddled with frameshift errors.  No good.

Other companies have tried to correct this by using error-correction enzymes that chew apart mismatched DNA to eliminate the random deletions.  While this is fruitful when using traditional plate oligos, the errors are so bad in microarray-sourced DNA that you end up with horrible final yields, and even if you do end up getting something correct in the end it takes so much additional screening work that the cost of post-processing erases any marginal gains made by beginning at the larger microarray scale.  I know several companies and academic groups that have investigated these batch pcr-strategies for making synthetic fragments from arrays, but believe that none have demonstrated sufficient speed and throughput of error-free synthetic fragments to make a convincing leap over traditional bulk-oligo assembly pipelines. (i.e. 15cents/bp for 1kb fragments from geneart w. 6 day turnaround).

We're basically building a lot of custom optical / sequencing / laser gear to sample single molecules of a microarray pool and to cherry-pick out the correct ones.  Instead of enzymatic tricks, we're doing physical separation upfront of the good from the bad, while simultaneously de-scrambling the ~10^5-10^6 pools into defined small assembly sets for very rapid, very high-throughput assembly of larger kilobase pieces.

The key spirit of what we're doing is to minimize biological complexity and trickery throughout the process and to backload complexity onto hardware and software - the domains where we're good at engineering.  I've been doing bio for over a decade, and if you don't keep it simple you're hosed!

-a

[Josiah Zayner]

I am interested in why the process is cheaper? Is it because you use less solutions?
I watched the video with Austen and comparing Sequencing to Synthesis seems like a moot point. Sequencing will always be able to be done cheaper because you can remove things from the reactions. Eventually, if people use carbon nanotubes the price of sequencing with be the price of just using the machine.
Whereas with Synthesis you are actively creating a quantity of something so your price can only be as cheap as the price of ligating together dNTPs. Even PCR probably costs $0.001 a base(enzyme($50/100 reactions, dNTP $100/200 reactions, buffers?? say you amplify 1000 bases). Throw in primer synthesis and template purification and it is closer to $0.01. Futhermore, doing it for $0.0001 or less is much different from selling it for that. Taking into account the costs of equipment, wo-manpower. I don't see DNA synthesis drastically dropping in price because it is reagent limiting.

"The price of gene synthesis has fallen drastically over the last decade. However, the current commercial price of gene synthesis, ~$0.40–1.00/bp, has begun to approach the relatively stable cost of the CPG oligonucleotide precursors (~$0.10–0.20/bp)¹, suggesting that oligonucleotide cost is limiting." Kosuri, Church and company Nature Biotech 2010

Using lasers seems cool but how much time does it save over taking a few minutes to run something over an HPLC column?

How are you assembling the kilobase pieces? Ligation? PCR? This seems like the slow step not the amplification part.

The method you are talking of is similar to one published by George Church and he says the error rate is only decreased 500 fold with the major errors being in wrong calls for sequencing and polymerase errors Cambrian Genomics method doesn't seem to decrease this? He mentions the error rate is still at 29% compared to what you said 36%. Matzas, Church and company Nature Biotech 2010

Thanks,
     Josiah Zayner

http://DoItOurselfScience.blogspot.com

[Anselm Levskaya]

Where to begin...

The cost of synthesis is a sum of the cost of the instruments (or its depreciation over time), the cost of the people running it, the fixed overhead costs of the company doing it, the cost of shipping, and lastly the cost of the reagents.  Among other things.  "Cost" is not a simple thing.

The cost of phosphoramidites is really low (at a minimum) given that they're produced in china.  The container the phosphoramidites come in can be more expensive than the contents.  CpG isn't that expensive either...  I mean let's take it to a ridiculous limit-point: an oligo synthesis run costs a few hundred dollars in reagents, maybe a thousand.  You're working with a ~mole of material.  Divide that by avogadro's number.  1e3/6e23 = about a tenth of an attocent per monomer.

If you had a magic daemon that coud assemble single-molecules, your reagent cost for a 100mer would be about 10 attocents.  (We don't know how to make magic daemons yet, but I suspect that physics does allow for a mesoscopically-coupled polymerase that could act as a decently accurate 'molecular printer' if you coupled it to some single-molecule sequencing feedback.)  Why does a bulk reagent limit the ultimate cost?  If you're efficient it doesn't have to.

Reagents are expensive when you waste them.  For traditional gene synthesis, the cost of oligos does dominate.  Even the lowest-scale bulk synthesis reactions make about 1000x more material than is needed for a synthesis reaction.  Microarray synthesis scales usage way down (though the amount of dna on a microarray spot is actually too low to use directly for reliable assembly, so some amount of post-processing is required)  However, microarrays jumble everything together and make the error-correction problem much more severe.

With regards to enzymes/reagents don't think that the price you pay for them is going to be the price an industrial purchaser is going to pay- most endusers of anything in biology are paying large margins.  What you're paying for as an enduser isn't the stuff in the tube, it's the quality control that guarantees the stuff in the tube will do what you hope it will, as well as the convenience of that little tube shipped to you the moment you need it.

Don't get fixated on a single aspect of the cost structure.  Ingots of ultrapure silicon aren't cheap, but their cost is only distantly related to the price of a cpu.  If you have a relatively fixed amount of cost for people and machines and reagents, the only way you get improved cost per unit is by making a shitload more units by using more efficient processes.  That's the idea.

>Using lasers seems cool but how much time does it save over taking a few minutes to run something over an HPLC column?

Have you ever run an HPLC column?  How do you propose to do it for a million different sequences?  You can't just run the whole batch: 1) there's not anywhere near enough material, and 2) HPLC retention times are highly sequence-dependent, not just length dependent, so it's not a batch purification strategy.

The paper you reference used a poor sequencing strategy that made exactly the same kinds of base-calling errors as the synthesis errors is was supposed to detect.  If you don't match your filter to your noise source, you're not going to get good rejection ratios.

Quick reference for error sources in DNA:

chemical synthesis errors : ~1 in 100

thermal depurination errors : 1 in 10^5 (or worse, depending on time spent above 60C)
polymerase copy errors: 1 in 10^6 (for modern, phusion-like enzymes)

in-vivo bacterial copy errors: 1 in 10^8-10^9 per replication.

Even if you magic away all the synthesis errors, you still have to worry about thermal depurination and large assembly errors.  It's all about building quality control into the whole system.  And that is why cheap sequencing is a big deal: it was a necessary precedent for cheap synthesis.  Of course, the devil is in the details of how you leverage it.

-A

[Nathan McCorkle]

I wonder what folks in the Caruther's lab are doing... seems like you
can only squeeze so much out of this control loop the way it's setup
now, before you simply have to go back to the basics and recajigger
the whole system.

http://chem.colorado.edu/index.php?option=com_content&view=article&id=246&Itemid=185