Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Biotech / Medical : momo-T/FIF -- Ignore unavailable to you. Want to Upgrade?

To: dr.praveen who wrote (4108)	3/16/2006 3:23:51 PM
From: dr.praveen	Read Replies (1) \| Respond to of 12215

Solexa/I have collected various Counter arguements from various boards for my post and I am posting them under one: (Sorry about the long post...Dr.Praveen) I love a good sequencing discussion. First my bias, I am a former ABI guy and have worked with and very much respect some members of 454's upper management. I have never worked in sequencing per se but follow the progress in the field. I hope you caught the article on 'next' generation sequencing in the January Scientific American. Unfortunately only the opening of the article is available online (free) sciam.com This article from The Scientist should also be of interest (I think it requires free registration). the-scientist.com And this older one from Genetic Engineering News (free reg) genengnews.com 454 is a majority owned subsidiary of CuraGen, which is public, but my understanding is that it is run as a separate entity. I would consider 454 a sequencing pure-play, but one that can only be participated in peripherally by buying Curagen. 454 is actively selling instruments right now and well ahead of Solexa in the market. This market is saturated with ABI capillary sequencers. Since the 'finish' of the genome project, sequencing has been a slowly diminishing market. It can show growth for a small company capturing market share (Solexa, 454, others), but unless the cheaper technology opens new markets such growth may well be limited. The challenge since the genome became available is and will continue to be determining relevance from the sequencing data. This 'relevance' is coming, and isn't all in genes or coding regions. Certainly broader coverage of the genome (more iterations) are required to find these associations, and that will require more and cheaper sequencing - but (captain Kirk voice) for how long. As sequencing comes down in price, less thorough technologies like real-time PCR and microarrays are advancing into providing coverage of sequences of known relevance. A $1000 dollar genome may sound great - but if the 'whole' genome hasn't shown relevance, and the relevant sections can be covered for $250, few will care (certainly not the insurance company paying for the clinical test). Celera's (current - LOL) business is to be based on using RT-PCR (maybe MS down the road) in clinical diagnostic settings, not de-novo sequencing. amazon.com Look to Cephied to expand offerings in this area as well. Affymetrix has already launched an array system to look at genetic variation in the p450 genes, thus rapidly assessing potential drug metabolism issues. I would expect more offerings from them along these lines as well. So, in my opinion, cheaper sequencing is great, and will be used - but isn't an invasive technology that will open new markets - rather it is a market that will be invaded. SLXA cannot assemble a human genome de novo using their methods. The limitation is the size of the read. The length of read is limited by the slowness of their methodology and the physical limitation of the genomic fragment attached to the plate. If the genomic fragment is much longer than 100 nt the noise between the individual fragments (represented by each colored dot on the plate) is too great and the process does not work. While I love the elegance and simplicity of their methods, without some type of new innovation it is not going to be used for any human genome project. 454's methodology is not as elegant, but it works and their is ample room for improvement. At this point, they are winning As someone who mouth-pipetted Sanger-style sequencing reactions into individual lanes, (back in the old days just before canostoga wagons went out of fad) I think this is up my alley. As far as the technology it is six of one; a half dozen of the other. Each technology has their advantage and their disadvantage. However, what makes me most hesistant is the number of nucleotides that can be obtained from each individual piece of genome in SLXA's method. Currently they have shown that 25 nts is >99% accuracy. At 50 nts the accuracy is dropping to 65%. All of the current alogrithims used to align DNA sequence rely on 30 nt overlap between fragments. In order to sequence and align all of the DNA required to complete an entire human genome, new software will be needed. Personally, I don't even know if it is possible to correctly align a human genome in 25 nt fragments. Right now, the ABI method depends upon 700 nt contiguous bps to complete a genome. Crgn with their next generation, which by the way was discussed in the addendum of the Oct Nature article, is able to achieve from the second generation machine 400 nt at >99% accuracy. So, we have a company (CRGN) that has published results showing sequencing results that can be used to align an entire human genome (although they have yet to do that). And we have a company (SLXA) that tells us it is going to produce a machine in the near future that can give results that will allow us to do this. That old adage about a bird in the hand is worth two in the bush keeps popping into my mind. CRGN has published and released all of the details regarding their technology. I can read it and see where the problems reside and where improvements in the technology will allow for advances. SLXA has not released much data, but from what I can see it is difficult to tell where process improvements will increase their speed and accuracy. one more detail. The whole argument about only exposing the nascent DNA chain to one nt at a time will introduce errors is moot. In the nature article, CRGN published and showed that the error rate due to misincorpation was 0.5%. Given that each bead holds 1000 copies of a single DNA fragment, the misincorporation rate is null. CRGN's biggest downfall is in its decreased accuracy when dealing with homopolymers >8 nt in length. The fact that 454 already has machines in place at sequencing centers , and is soon to launch the next generation is important. Essentially, they have delivered on their promises. The company claims that the second generation machine can read out to 400 nt from each well with the same accuracy -- 96%. Slxa while intriguing is still in development. They have only shown that their system can sequence a <200,000 nt artificial human chromosome. We wait for confirmation of their claims. 454 will always beat SLXA in pure speed because there is only 1 enzymatic/chemical step for each nt addition. The design of the SLXA process is better able to handle homopolymers. Assembling a human genome in 25 nt chunks is a daunting task, and I have to say that I am skeptical. Increasing the capability of each read to 50 bp will improve the method, however, the accuracy, which is the main feature of the SLXA system, is then lost. I await judgement until results of the system for sequencing a larger genome are released. In closing.... back in the old days two methods of sequencing were invented, the Sanger method which relied on enzymes and Maxim/Gilbert which relied on a chemical reaction. Maxim/Gilbert was more accurate, but the "runs" were shorter. Sanger had the advantage of being more flexible and given to improvements for speed and distance. The same story is building here with Solexa and 454; and while I can't predict who the winner will be, I don't see Solexa as the front runner at this time..