backtop


Print 15 comment(s) - last by jimhsu.. on Jun 29 at 11:25 AM

Scientists thing the new method can save millions in sequencing costs

Scientists at the Cold Spring Harbor Laboratory (CSHL) have discovered a new method of pooling a multitude of DNA samples for sequencing. The new method is being called DNA Sudoku because it uses a method similar to the math game to greatly increase the speed at which DNA can be sequenced.

The researchers report that DNA Sudoku method allows for tens of thousands of DNA samples, short combinations of polynucleotides with A, T, G, and C bases, to be combined and sequenced.  The simultaneous sequencing is done by looking at the letter order and comparing it to the correct order of the known human genome using an algorithm that resembles those used to solve Soduku puzzles.

The ability to do all the sequencing at once is a massive improvement over past methods that allowed only a single DNA sample to be sequenced at a time. It is also an improvement on current techniques that can ideally only combine hundreds of samples.

Researcher Gregory Hannon, Ph.D. said, "In theory, it is possible to use the Sudoku method to sequence more than a hundred thousand DNA samples." Hannon is a genomics expert and the leader of the team that invented the Sudoku approach.

The Sudoku method was originally developed to help solve a problem that was plaguing one of the labs research projects and the researchers discovered that the method had promise for clinical applications. According to Hannon, the technique can be used to analyze specific regions of genomes form a large population to identify those who carry genetic mutations that may cause genetic disease.

Another of the key benefits of the Sudoku technique is that it is much cheaper that the current methods used by researchers. A sequencing project using current methods can cast around $10 million according to Hannon and using his team's new approach a project of similar size might possibly be done for $50,000 to $80,000.

In traditional multiplexing -- sequencing high number of DNA samples at one -- each sample had to be first tagged with a barcode. With the Sudoku method the researchers would only need to tag each pool of samples with a barcode.

Researcher Yaniv Erlich said, "It's time-consuming and costly to have to design a unique barcode for each sample prior to sequencing, especially if the number of samples runs in the thousands."

Erlich continued saying, "Since we know which pool contains which samples [with the Sudoku method], we can link a sequence to an individual sample with high confidence."

The Sudoku method is currently ideally used on genotype analyses that need only short segments of an individual's genome. Clinical applications for the new method could be for HLA typing, which is a diagnostic tool for predicting the risk of organ transplant, cancer, and autoimmune disease.


Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Not quite as revolutionary as it seems
By zsunjian on 6/25/2009 12:29:10 PM , Rating: 2
It sounds quite interesting; I'll go check out their original paper. Surprisingly I haven't heard about this in our circle.
In any case the current cost of genome sequencing is not $10 million, as the guy in this post claims. Illumina claimed to have sequenced full genome under a month for only $100k back in Feb 2008. They are some other companies offering sequencing for even less - just google; but I personally have more confidence in the Solexa or Illumina platforms.
Now, Pair-End Sequencing is much more information-rich.




By zsunjian on 6/25/2009 12:36:01 PM , Rating: 2
What he means by a "sequencing project" can be important. Obviously if he's talking about sequencing 100 genomes as a project (which would be exceedingly ambitious and put the current method's tab at $10 mil) for about $50k-$80k, then kudos to their lab. But I very much doubt if that's the case.


RE: Not quite as revolutionary as it seems
By Screwballl on 6/25/2009 12:42:34 PM , Rating: 2
I know nothing about this and am just going to nod my head as if I know what you are talking about...

Although my question is WHY is this on a TECH website?


By zsunjian on 6/25/2009 12:58:26 PM , Rating: 5
I guess it's because we call it "biotech"...


By jimhsu on 6/29/2009 11:22:22 AM , Rating: 2
Some time ago a group developed a camera-like technology to actually photograph individual bases (probably w/ fluorescent tags attached). Something about giant reads (50-100K) which vastly exceeds anything else on the market. I wonder if that tech ever got developed.


Sudoku != math
By Kanji on 6/25/2009 7:42:36 PM , Rating: 2
quote:
The new method is being called DNA Sudoku because it uses a method similar to the math game


Sudoku is not a math game. I never do any sort of math to solve those puzzles. You can replace the numbers with any sort of symbol or picture and it works EXACTLY the same. Just because it has numbers does not make it a math puzzle. Why would anyone think it is a math puzzle??




RE: Sudoku != math
By Brain onna Bun on 6/26/2009 4:28:24 AM , Rating: 3
quote:
Sudoku is not a math game. I never do any sort of math to solve those puzzles.

Counting from 1-9 is a form of math, just extremely elementary math.

quote:
You can replace the numbers with any sort of symbol or picture and it works EXACTLY the same.

Algebra is a form of math where the numerical values are replaced by placeholders, letters usually, but the placeholders can be of any kind.


RE: Sudoku != math
By FormulaRedline on 6/26/2009 10:16:37 AM , Rating: 3
No, Kanji is completely right. There is absolutely no math involved in this game.

For example, you could easily replace the numbers with farm animals. Then, to solve a certain row, I don't need to count the animals, I only need to see that it is missing a chicken and a pig.

Logic? Process of elimination? Yes. Math? No.


RE: Sudoku != math
By shecknoscopy on 6/26/2009 11:37:08 AM , Rating: 4
quote:
No, Kanji is completely right. There is absolutely no math involved in this game.
[...]Logic? Process of elimination? Yes. Math? No.


Whoa whoo, hold the phone! There's no arithmetic involved in solving a Sudoku to be sure, but - as you yourself concede, there's a tremendous amount of logic involved. Logic is a branch of mathematics (see Bertrand Russel, Kurt Godel &ct...), and forms the core of math's most basic component: the proof. The process of completing a Sudoku is really just a series of proofs.

Thrm That little square should be a "3."
Proof Numbers 1,2, 4-9 have been placed in that block of nine, leaving only 3 remaining.
Lemma There are two other 3's already in that column.
QED: I suck at Sudoku.

Anyway, you get the picture. Tomorrow, I'll (controversially) argue that there's no math in Ken-Ken. Only frustration.

-Sheq


The Emergence of "BioBanks"
By Shig on 6/25/2009 2:08:19 PM , Rating: 2
http://en.wikipedia.org/wiki/Biobanks

In Europe they are beginning to create banks that pool all of that genetic information in one central location for a given population. From all of that data they try to cross-examine a large population's genetic information with their entire medical history. They hope this will lead them to detect diseases years earlier than was ever possible before, or maybe tell someone they are genetically prone to a certain ailment and help that person take preventative measures before anything can even go wrong in the first place.

Ethics and Insurance Policies are two key areas in which we have a lot of work to do still with genetics. A lot of people would not like insurance companies discriminating against them and raising their rates simply based on their genetic abnormalities. Personally I wouldn't mind donating my genetic information to a bank in hopes it could help others, but a lot of others would not share my point of view and wouldn't want strangers playing with their genetic information.

Looking into the future. The cost of sequencing a human's genome will be so cheap in 10-20 years, it will almost be trivial and probably part of a normal checkup at the doctor's office.




RE: The Emergence of "BioBanks"
By PigLickJF on 6/25/2009 3:59:00 PM , Rating: 2
I don't think it will be something they'll do at every visit, as it's something that should only need to be done once.

I can see it being added to/replacing the newborn screening tests we already use, and then becoming part of that person's medical record.

Of course, that's ignoring the political/social/economics aspects mentioned above in terms of insurance, etc. Those issues could easily prevent widespread, routine genotyping from becoming the norm anytime soon, no matter how simple/cheap the process itself becomes.


RE: The Emergence of "BioBanks"
By zsunjian on 6/25/2009 5:15:23 PM , Rating: 2
Yes and no. I can see that in the distant future one might get a DNA sequencing done as baseline reading. Then if one gets a tumor, which often involves significant mutations (single SNP mutation, CNV, ins/del events, translocation), they could get the tumor biopsy and sequence the tumor cells for personalized cancer treatment.
The use of genetics will obviously lead to ethical concerns (Gattaca, anyone?); but we are still very far away from actually understanding what your genome is saying. And of course your life isn't dictated by genetics, either.


RE: The Emergence of "BioBanks"
By jimhsu on 6/29/2009 11:25:52 AM , Rating: 2
Ah GATTACA, I remember that movie.

Another thing likely, if Affy and co. can lower their prices, is to do gene expression analysis of patients, and see if certain markers (e.g. metastasis) are over-under/expressed. In fact that might be better, since it can be that problems aren't necessarily caused by mutations at the DNA level.


error error!
By xDrift0rx on 6/25/2009 12:46:12 PM , Rating: 2
"Scientists thing the new method can save millions in sequencing costs"

i think you meant "Scientists think the new method can save millions in sequencing costs"




Proofreading?
By concerned1 on 6/26/2009 3:46:15 AM , Rating: 2
Doesn't anyone PROOFREAD these articles before they are published? I found numerous misspellings and other errors that are glaringly wrong. Obviously, the writer/editor is relying on the spell-checker to do his/her proofreading. None of the misspellings were in themselves wrongly spelled words, but in at least two cases, they weren't the words that were meant. By not accurately proofreading your articles before publishing them, you only prove that you don't care about how the readers think about you as a writer.




"If they're going to pirate somebody, we want it to be us rather than somebody else." -- Microsoft Business Group President Jeff Raikes











botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki