Tweet
Hi folks!
I registered here after reading a Keith Robinson's blog on Oxford nanopore technology.
His last paragraph is:
A last thought: for any company considering doing this, please <snip> Do yourself a favor: build a registration-free data release site and just use SEQAnswers as the discussion forum. Your marketing people will grouse you've missed an opportunity to collect data, but just ignore them. You will have something far more precious: lots of ADHDish programmers trying to play with your platform's data.
"Lots of ADHDish programmers" sounded like a place I might want to visit, so I've stopped by.
I'm principally interested in genome annotation for metabolic modeling, and have published on those subjects. More recently, I stepped into the world of metatranscriptomics and metagenomics for environmental samples of microbial communities, and really think that short reads (less than a typical prokaryotic gene or 1000bp) are the wrong tools for this particular job. I think I'll hold out plunging in deeper until someone (Oxford?) can come up with reads longer than 1000bp that are 99.99% accurate. In other words, at least as good as ancient ABI machines, but faster. PacBio at 3kbp and 85% doesn't cut it with me. Oxford at 30kbp and 96% accuracy doesn't cut it for me. "Slow down and reread for error correction" is what both companies propose for their dismal error rates compared to the slow but accurate first generation machines they replace is perhaps a solution. But how much slower and how much better? I'm waiting. Like Keith Robinson, I want to see hard data, realistic data, not hyped projections and fake data tuning to immature technology.
The two domains of life making up the prokaryotes (sorry Norm Pace, it's a useful word in my opinion) are far too diverse to be studied fruitfully with current NextGen technology, again in my opinion. That technology is awesome for resequencing to a reference, e.g. human genomes for medical records. It is a claimed solution to the problem of metagenomics and metatranscriptomics on the theory that depth of sequencing can cover up the inherent problems of 150bp reads from the prokaryotes. Our choked pipelines at sequencing centers bear me out, I think.
Nice chatting. Cheers to all.
I registered here after reading a Keith Robinson's blog on Oxford nanopore technology.
His last paragraph is:
A last thought: for any company considering doing this, please <snip> Do yourself a favor: build a registration-free data release site and just use SEQAnswers as the discussion forum. Your marketing people will grouse you've missed an opportunity to collect data, but just ignore them. You will have something far more precious: lots of ADHDish programmers trying to play with your platform's data.
"Lots of ADHDish programmers" sounded like a place I might want to visit, so I've stopped by.
I'm principally interested in genome annotation for metabolic modeling, and have published on those subjects. More recently, I stepped into the world of metatranscriptomics and metagenomics for environmental samples of microbial communities, and really think that short reads (less than a typical prokaryotic gene or 1000bp) are the wrong tools for this particular job. I think I'll hold out plunging in deeper until someone (Oxford?) can come up with reads longer than 1000bp that are 99.99% accurate. In other words, at least as good as ancient ABI machines, but faster. PacBio at 3kbp and 85% doesn't cut it with me. Oxford at 30kbp and 96% accuracy doesn't cut it for me. "Slow down and reread for error correction" is what both companies propose for their dismal error rates compared to the slow but accurate first generation machines they replace is perhaps a solution. But how much slower and how much better? I'm waiting. Like Keith Robinson, I want to see hard data, realistic data, not hyped projections and fake data tuning to immature technology.
The two domains of life making up the prokaryotes (sorry Norm Pace, it's a useful word in my opinion) are far too diverse to be studied fruitfully with current NextGen technology, again in my opinion. That technology is awesome for resequencing to a reference, e.g. human genomes for medical records. It is a claimed solution to the problem of metagenomics and metatranscriptomics on the theory that depth of sequencing can cover up the inherent problems of 150bp reads from the prokaryotes. Our choked pipelines at sequencing centers bear me out, I think.
Nice chatting. Cheers to all.
Comment