Real time of the whole process

[joscarhuguet]

I have been doing some research about prices and time for sequencing using any of the new technologies (ilumina/solexa, 454). Although in theory the sequencing reaction and all the additional chemistry in the lab could happen very fast....less than one week. The big bottle neck is the time how samples are hold until lanes or complete sequencing-chips are complete. In my specific case I just wanted to do one lane of paired-end 2x 100 bp ilumina. In theory should be fast. However consulting with several facilities I calculate a time of 6 to 9 weeks!..... Or if you want it faster you should use all the lanes.. so is not as cheap as they advertise. That is not the price and time that you hear when to go to all these new presentations that say that sequencing now is a "commodity".
Perhaps I am missing something and there are some facilities that take less time. Any comments....

[Bukowski]

Just remember that when you send something to be outsourced, there's a lot more to it than just putting it on a lane and hitting go. I think an 6-8 week turn around time is pretty good. In practice you will find this a lot longer for some companies, so if you find a place which someone has used before and you can get them to honour an 6-8 week TAT - feel good about things.

If you're doing a SureSelect capture the process takes days.

An Illumina HiSeq 2000 run takes *11 days*.

When you send your DNA to a service provider they don't just process it, it will be QC'd on a Nanodrop or Qbit, every step of the way will be QC'd. It's time consuming.

If there's bioinformatics services, then they will take their time as well - do you need the data aligned? Variant called? It all takes time.

Then there is the time that is factored in, in case something goes wrong. Let's say your HiSeq run fails and you have to repeat it. At some point in the 6-8 week figure, that will be factored in too - companies need leeway to protect against machine failure etc.

If you want the advertised prices, then you have to run under the agreement. Yes there is a wait to perhaps fill a lane/flowcell if small sample numbers are coming in - but this is so machines can be run at capacity. If the machines run to capacity, they run most cost-efficiently and it is you, the consumer, who benefits from that in the end.

[joscarhuguet]

agree ...so genome sequencing is not a "trivial lab-thing" yet, and scientific journals should give more credit for genome publications.

[huguesparri]

And that's also why the 2500 upgrade can be very interesting: it can shorten drastically the wait time due to flow cell completion.