Reduction in cost of DNA analysis

[CharonZ]

Actually it is not possible to measure the bases directly with an AFM tip.
Except for measuring currents maybe, however this can be easily realized by electrodes to (and the setup would be easier, too).

 

[ElectricFetus]

Actually it is not possible to measure the bases directly with an AFM tip.
Except for measuring currents maybe, however this can be easily realized by electrodes to (and the setup would be easier, too).

Yes but the tip has to be sharp to the atom, as a result your going to get electron jumping so its will be sort of like a AFM. I would imagine the tips will have functional groups attached with different binding affinity to different base pairs.

To be clear we are just exploring hypothetical here, researchers who are actively prosing this technology are not giving up patentable details so were a just conjecturing how they are going to get over certain problems.

 

[Billy T]

I remember a photo (I forget how it was observed - doubt it was from some super electrong mictoscope?) of a gold alloy wire being slowly pulled apart. I forget details but strange thing was that although the gold was only about 10% of the alloy (I forget what the rest was.) the gold was the only type atoms in the single atom line just before it broke. Thus it is possible to make a needle tip of gold only one atom in diameter, perhaps even a "spike cylinder" or short wisker of two or three gold atoms in a line. I am sure that the chemical properties would differ from bulk gold but that would seem to be a gold thing to try for an electrode that might some how explore the conductivity of the cross links of DNA.

Yes, I knew that one put the cleaved pieces of DNA together in a computer. In fact I think that computer technology was first well developed by Graig Ventner to scoop the US effort (but being smart about where the money might be coming from, he did not use the word "scoop" etc.)

As far as how one might get straight pieces I had in mind some way (electric field?) to align them in a liquid (ETOH?) that would evaporate, but leave DNA ends streched across some less volital slightly curved frame. Sort of like the sting of a bow in a bow&arrow - just wild ill-defined ideas I know, but it should not be very hard to make a solid piece with "bows" by the 10s of thousands, close spaced side-by-side, and to have a few DNA strands end up as "DNA strings" across them (or between oppsite ends of near by "bows").

 

[Billy T]

This thread started out in large part about Ilumina and the rapid advances being made in sequencing. Below in bold part of text this is made very clear:

"... David B. Goldstein, Ph.D., Director of the Center for Human Genome Variation at Duke’s Institute for Genome Sciences and Policy will lead the investigation to help explain why some people are able to fend off infection even when repeatedly exposed to HIV. “This study represents the first effort to search through the entire human genome for any variants, rare or common, that influence susceptibility to infection by HIV-1. Working with Illumina’s next-generation sequencing technologies will allow us to rapidly and accurately screen the genomes of 50 individuals at very high coverage, amounting to the generation of as much data as was generated by the human genome project every two to three weeks. These studies will allow us to identify the most important genetic determinants of resistance to HIV-1 and may ultimately point the way toward new therapies to protect against HIV/AIDS,” said Goldstein.

“The decision by Duke University Medical Center to use Genome Analyzers is additional validation that our sequencing platform is delivering leading performance with its unmatched rate of daily output, ease of use, and proven paired-end sequencing capabilities,” said Joel McComb, Senior Vice-President and General Manager of Illumina’s Life Sciences Business Unit. ..."

You can disreguard the last self promotional paragraph - included on to show the source. Duke made their decision and did chose Ilummina's "cutting edge" machines /technology. "The Duke University research project is supported by a $3 million grant from the Bill and Melinda Gates Foundation."
read full report at:
http://www.freshnews.com/news/biotech-biomedical/article_47371.html

 

[Roman]

What extraction protocol do you use for doing whole genomes?

 

[Billy T]

What extraction protocol do you use for doing whole genomes?

I do not know. Do not understand much of this sequencing, snips etc. Perhaps you can learn at their site: www.illumina.com

 

[CharonZ]

What extraction protocol do you use for doing whole genomes?

Actually standard protocols. At some point you shear them to build libraries anyway. I recall that for the next-gen sequencers you need a higher purity then you get with a number of the kits though. Some to density centrifugation as additional step (IIRC).

Working with Illumina’s next-generation sequencing technologies will allow us to rapidly and accurately screen the genomes of 50 individuals at very high coverage, amounting to the generation of as much data as was generated by the human genome project every two to three weeks.

This is some ingenious wording. They get the amount of sequence as in the HUGO project, but due to the relatively short read length (they are increasing but are still behind 454) they will not have the complete genome. Hence they say screening of genomes and not sequencing. But bringing up the HUGO project they associate themselves with it. Clever. I wish I could think of writing that way.

In any case, a colleague of mine (who has got a 454) might be getting a solexa eventually. They are going to use the 454 for sequencing and the solexa for HapMapping and ChiP analyses. I am really interested whether the data does really boost research.

 

[Billy T]

ILMN announced today {6Feb09} that it has reached an agreement in principle for the Genome Center at Washington University School of Medicine in St. Louis to acquire 21 genome analyzers over the next several months to support the Center's expanding research initiatives. Once the scale-up is complete, the Center will have an installed base of 35 Genome Analyzers. The added capacity, along with the continued improvements to the Genome Analyzer platform, will allow the Genome Center to sequence in the order of one human genome per day at 25x coverage. *

"Our intention to substantially scale-up with this technology reflects our commitment to large-scale sequencing projects that aim to uncover the underlying genetic basis of various human diseases. With the rapid decline in the cost of whole-genome sequencing, we believe now is the time to embark on initiatives which were previously not possible," said Richard K. Wilson, Ph.D., Professor of Genetics and Director of the Genome Center at Washington University. ...

The Genome Center at Washington University is a world leader in genomics research. {Is that true or just self promotion?} The Center played a leadership role in The Human Genome Project and currently is a leading participant in both The Cancer Genome Atlas (TCGA: 69.4, 0.79, 1.15%) project and The 1000 Genomes Project.

FROM: http://www.foxbusiness.com/story/ma...ington-university-scales-illumina-sequencers/
------------
*What does 25x coverage mean? Based on Charon's comment in post 87 I am guessing that is the "read length" (Number of base pairs in the chains they identify?)

 

[CharonZ]

Re: coverage. As the genome is never sequenced in one go it has to be assembled by finding overlaps in the sequenced fragments. However as you cannot control too much what fragment you get sequenced (hence the name shotgun-sequencing) you need to sequence a total that is higher than your actual genome size to ensure that you got everything.
This also depends on the read length (of which sanger sequencing still have the longest, then followed by 454 and and Solexa having the lowest to date).

So a coverage of 25x indicates that you got a total of sequences that equal 25 times the whole genome length. This might be not enough for a human genome (which has an abundance of repeated elements) using short-read machines.

 

[Billy T]

...So a coverage of 25x indicates that you got a total of sequences that equal 25 times the whole genome length. This might be not enough for a human genome (which has an abundance of repeated elements) using short-read machines.

Thanks. I had guessed wrong what "25x coverage" might be.

 

[CharonZ]

Sure thing. This is something that derived from lab-lingo and is not necessarily easily understandable.
On a different note, I cam across this http://www.completegenomics.com/pages/materials/CompleteGenomicsPOC_PressRelease.pdf

Essentially they claimed to have a sequence of the human genome with 91x coverage. It is also using short reads only, though. The interesting bit is that they claim (somewhere else) to be able to do the sequencing for around 5000$ a shot. I am not sure whether it is for a single run or for the 91x coverage (which took nine runs in total).

 

[Billy T]

...I cam across this ...

Interesting. Thanks again.

They seem to be privately held, by venture capital and founders. Their "bead array" sequencers is probably different from ILMN's but as only two years old, they may still have some patten over lap problems with ILMN. They appear to be planing on selling sequencing service (and doing R&D) not selling machines to those who do.

Their home page is: [URL]http://www.completegenomics.com/ but not very informative -still building, I think, as some headers have nothing when you pull down.
I will visit occasionally. May even decide to take profits in ILMN as I will have plenty of offsetting tax loses I can take.

Hear anything more of interest like this one, let me (and others here) know.

 

[Billy T]

I ill-understand the following, but it sounds impressive and the stock has more than doubled since I first started posting about their analysis methods here. (Considering the general decline of stocks in the same period - that too is impressive.)

"... Illumina, Inc. (NASDAQ:ILMN) today announced a new product for DNA Analysis: the HumanOmni1-Quad BeadChip. With over four million data points on a single BeadChip, this array includes up-to-date content for all major classes of genetic variation for the study of human disease. The BeadChip utilizes tagSNPs from all three phases of the International HapMap Project and offers premier coverage of known regions of copy number variation (CNV) and SNPs of known disease association. In addition, the HumanOmni1-Quad BeadChip is the first commercially-available product to offer content derived from the 1,000 Genomes Project. The format of the HumanOmni1-Quad BeadChip supports simultaneous genotyping of four samples which increases throughput and decreases the per sample price. First customer shipments of HumanOmni1-Quad BeadChips are expected in June 2009.

The HumanOmni1-Quad BeadChip targets more than 11,000 regions of both common and rare copy number variation developed in collaboration with The Centre for Applied Genomics at the Hospital for Sick Children in Toronto, the Wellcome Trust Sanger Institute in the United Kingdom and Harvard Medical School / Brigham and Women's Hospital. BeadChip content also includes over 30,000 non-synonymous SNPs, indels and additional functional variants not found on other arrays. The HumanOmni1-Quad BeadChip delivers carefully selected markers that capture the highest amount of genomic variation and the best power for genome-wide association studies (GWAS) and CNV discovery and screening. ..."

More details at: http://news.morningstar.com/newsnet/ViewNews.aspx?article=/BW/20090518005498_univ.xml or ILMN's web site, I assume.

 

[Billy T]

This MAY be a giant step to low cost sequencing:
http://www.forbes.com/forbes/2011/0...ch-gene-machine.html?partner=daily_newsletter

How does top half's letter disappear?

 

[Hercules Rockefeller]

How does top half's letter disappear?

You start with single-stranded DNA and use a DNA polymerase enzyme to create a complementary strand (ie. create double-stranded DNA) from free nucleotides.

As you say, this may have an effect on the genomics field, or it may not. I guess accuracy is an issue. Fast isn't always good if the resulting sequences are inaccurate. It all depends on the application and the questions being asked.

-----------------------
Now, I read the first page of that Forbes article (only the first page, I didn’t have time to read all 6 pages), and I have to comment on a particular paragraph that, for the most part, is complete rubbish.

You think you've heard this before, don't you? Genomics has certainly been overhyped--and so far failed to deliver on its promises. Many intelligent people have relegated the idea to the dusty corner shared by hopes for cold fusion, world peace and World Series rings for the Chicago Cubs. When scientists first mapped the human genome a decade ago, they bragged it would lead to cures for Alzheimer's, heart disease, schizophrenia and more. It hasn't happened. Drug approvals have gone down. The search for the genetic roots of heart disease, diabetes and other common ills has yielded surprisingly little useful information for the average person. Even 23andMe, the high-profile consumer gene-testing company cofounded by Anne Wojcicki, the wife of Google's Sergey Brin, had to lay off people last year.

What a load of twaddle! Let’s break it down a bit:

Genomics has certainly been overhyped--and so far failed to deliver on its promises.

Utter and complete rubbish. Genomics is a well-established field of research that provides numerous valuable insights into both basic biology (genetics, biochemistry, development, evolution) and medical research (genetics disorders, developmental disorders, natural variation that underpins disease susceptibility and differences in drug response, cancer). There are numerous dedicated journals for the field and genomic research frequently appears in the top generalist journals (Science, Nature etc.)

Many intelligent people have relegated the idea to the dusty corner shared by hopes for cold fusion, world peace and World Series rings for the Chicago Cubs.

Well, clearly the author is not one of these “intelligent people” of which he speaks.

When scientists first mapped the human genome a decade ago, they bragged it would lead to cures for Alzheimer's, heart disease, schizophrenia and more.

They did not. No reasonable scientist who knows anything of this area would have made such predictions. These sorts of ridiculous claims were made by journalists like the author. It was the media who are responsible for over-hyping, not scientists performing such research.

It hasn't happened. Drug approvals have gone down. The search for the genetic roots of heart disease, diabetes and other common ills has yielded surprisingly little useful information for the average person.

More ill-informed commentary. Research into heart disease, diabetes and other common ills has steadily progressed over the last decade. We know more about these diseases now than we did in 2000 when the first draft human genome sequences were published. And if drug approvals are down, it’s because we demand so much from new drugs. The efficacy and specificity that we require of new drugs continues to increase. I collaborate with a scientist who is in the drug discovery field. I have learnt from him just how difficult it is to find/create compounds that fulfil these requirements.

Even 23andMe, the high-profile consumer gene-testing company cofounded by Anne Wojcicki, the wife of Google's Sergey Brin, had to lay off people last year.

Consumer services such as this has nothing to do with genomics. This is simple genetic testing, ie. determination of the presence or absence of several already-known alleles of known genes. This falls well short of anything that would be considered as genomics, ie. looking at whole genomes.

 

[Billy T]

... I collaborate with a scientist who is in the drug discovery field. I have learnt from him just how difficult it is to find/create compounds that fulfil these requirements. ...

I completely agree.

On 10Jan07 while listening via the internet to a presentation by AtheroGenics I learned that the abstract was wrong. Too few of the clinical trial patients had died. – No results were yet available. The speaker apologized and told what he could. I forget the numbers, but they were impressive. Sort of like 80 of these serious heart problem patients should have died but only 55 had; So before the speaker had finished his talk, I bought 100sh of AGIX for $1094.99 but on 7Oct08, AGIX filed for Chapter 11 bankruptcy.

What went wrong, was not the drug, but the excessive claims AGIX made for it to the FDA. Their set of claims was long – at least five different ones and two or three were not meet. For example, the number of patients needing and receiving a heart transplant had actually increased - not the reduction AGIX had claimed. But this was because the drug had prevented them from dying for many months, etc. so they lived long enough to become stronger with the drug aiding them and got their heart transplant!

Needless to say I was upset by the FDA’s rigid application of its rules – AGIX failed to meet all its claims so drug was not approved. I think that now lost drug could have saved many lives, but it was not approved and the company stock sharply declined. This made it impossible for them to float more stock to raise cash to continue new data collection trial with a subsequent application to FDA for the drug that only the claimed that it extended lives of patients needing heart transplant long enough for them to get one and become stronger for better chance of surviving with a new heart.

I knew there was huge risk but then held stock in about 35 early stage drug developers and could afford the loss, but I really feel sad that this potentially important drug never got a reasonable / more modest / approval request sent to the FDA.

Main point being that even a good, safe drug may fail to make it to market.

 

[S.A.M.]

I knew there was huge risk but then held stock in about 35 early stage drug developers and could afford the loss, but I really feel sad that this potentially important drug never got a reasonable / more modest / approval request sent to the FDA.

The problem with FDA is not so much their rules or rigidity, its their politics.

The industry, funding over 80% of trials, sets up a research agenda guided more by marketing than by clinical considerations. Smart statistical and epidemiological tactics help obtain the desired results. Budget for marketing is by far greater than for research. Massive advertising to physicians and to the public gets increasingly sophisticated: ghost writing, professional guidelines, targeting of consumer groups and manipulating media for disease mongering. Pervasive lobbying and political ties limit the independence of regulatory bodies. Obligation to shareholders overriding public health considerations is not unique to the pharmaceutical industry. The chemical, tobacco and food industries share similar tactics: proclaiming doubts about safety issues, buying researchers, infiltrating universities, boards, media and legislative agencies. By contrast, powerful and cheap health promoting activities, poorly supported by industry because they are too cheap and not patented, are markedly underutilized: technologies for changing behavior (e.g., cardiac rehabilitation), palliative care and use of old, effective and safe drugs - all could benefit from industry's tools of marketing and quality. As those most affected are the sick, the poor and the least educated, free market successes appear to pose unsolvable challenges to social justice in public health.

http://www.ncbi.nlm.nih.gov/pubmed/18982834

So, its hardly surprising that a large number of appointees in the FDA [I believe the figure is around 90%] come from private pharma industry.

You may be interested in this:

Critical Condition: How Health Care in America Became Big Business--and Bad Medicine

 

[Hercules Rockefeller]

I can’t comment one way or the other on all that conspiracy theory regarding the FDA approval process as (i) I’m not American and (ii) I’m at the basic research end of the spectrum and not the drug development end.

However, I agree somewhat with this premise:

By contrast, powerful and cheap health promoting activities, poorly supported by industry because they are too cheap and not patented, are markedly underutilized: technologies for changing behavior (e.g., cardiac rehabilitation), palliative care and use of old, effective and safe drugs - all could benefit from industry's tools of marketing and quality.

Our consumer-driven capitalist Western societies seem to have produced a noticeable move away from personal medical responsibility (ie. behavioural modification) towards reliance on new drugs to fix medical complaints.

Anyway, this is veering away from the topic of DNA analysis.

 

[Billy T]

I can’t comment one way or the other on all that conspiracy theory regarding the FDA approval process...

Well I can and have so again briefly:
GWB enjoyed nearly 100% support from the relgious right wing and many others who were strongly opposed to abortions. He was careful to not anger them. When Barr Pharma (now owned by Teva)* developed and carefully tested their "morning after pill" cleverly called "Plan B" for commercial sales, more than one scientific review committee (FDA often forms these groups of experts) recommended approval, in one case by 100% majority. But GWB ordered the head of the FDA to delay more. For some years Plan B was not approved or had complex requirement (a doctor's prescription, parental approval for the under 21, and no availably even with parental approved for the under 18.)

Congress eventually got evolved, especially Senator Hillary Clinton, in support of Plan B, which had been recommended by the scientific committee as an "over the counter" drug for 18 year and older women in need. I forget how many days after unprotected sex the drug was still highly effective, but about the same as would be required to get an appointment with a doctor. Eventually the congressional pressure (and other factors?) forced the head of the FDA to resign and under the new one, Plan B was approved, still with the requirement that it not be available to the under 18 year olds (unless with doctor's prescription? - I forget). This was not a great limitation in practice as drug could be sold to a 19 year old male, without a prescription. (Law and FDA rules did not specify the required sex of the buyer!)

But as Herc notes, this is somewhat off thread’s subject; however, it is important to understand that politics does strongly enter the drug (or medical device) approval process. For example, when cheap DNA analysis is available and many more disease markers in the DNA are known , you can be sure some will want to limit those types of results to only doctors, not the person whose DNA was studied. This probably will reduce the demand and thus increase the cost of DNA analysis. (Doctors do charge fees, you know, so if forced to go thru one, you will pay more to learn your genetic risks.)

* I know about this history as I owned stock in both Barr & Teva, (only Teva now) which has been financially rewarding although Teva is off its highs as there soon may be generic versions of its biggest winner, a drug for MS and probably other new MS drugs that can be taken orally, instead of by injection.

The generic will be a "bio-similar" and there is a lot yet to be firmed up about how extensive the testing of bio-similar must be. Teva, of course, supports the POV that they must be fully tested like any new drug as that would delay the competitor for a few more years.

 

[Billy T]

“ … Jan. 10, 2012: Life Technologies Corporation today announced it is taking orders for the new benchtop Ion Proton™Sequencer that is designed to sequence the entire human genome in a day for $1,000. The Ion Proton™ Sequencer, priced at $149,000, is based on the next generation of semiconductor sequencing technology that has made its predecessor, the Ion Personal Genome Machine™ (PGM™), the fastest-selling sequencer in the world.

Up to now, it has taken weeks or months to sequence a human genome at a cost of $5,000 to $10,000 using optical-based sequencing technologies. The slow pace and the high instrument cost of $500,000 to $750,000 have limited human genome sequencing to relatively few research labs.

Baylor College of Medicine, Yale School of Medicine, and The Broad Institute, have each signed up for multiple Ion Proton™Sequencers and will be the first customers to adopt this transformative technology. …”

See it here:
(Photo: http://photos.prnewswire.com/prnh/20120110/LA31914-a) The "chip" heart of machine, I think
(Photo: http://photos.prnewswire.com/prnh/20120110/LA31914-b) The small machine.

From: http://www.prnewswire.com/news-rele...man-genome-in-one-day-for-1000-136990438.html