The 'cost' of sequencing

[Graham Etherington]

Hi all,
I'm currently writing an article about sequencing and am trying make fair comparisons between the cost of sequencing 1Mb of DNA on the following machines:

ABI 3730xl (Sanger sequencing)
ABI Solid4hq
Illumina HiSeq2000
Roche 454 GS FLX+

The only manufacturers figures that I can get is for the ABI Solid4hq, which is $33 per Mb (based on their claim of $3000 for sequencing of the human genome at 30X coverage). I'm presuming this includes library prep.

I was wondering if anyone could give me some information about the cost of sequencing on the other machines. Most information that I can find in literature searches is a couple of years old, so I don't feel that I can make a good comparison with current estimates. Also feel free to comment on ABI's estimate above.
Many thanks,
Graham

[ulz_peter]

That's a difficult one, especially if you want to include Sanger sequencing. As the sequencing information in sanger sequencing is derived of thousands of copies of the molecules and therefore is of high quality, 1MB with NGS is prone to sequencing artifacts. So 1MB Sanger sequencing will appear too high in comparison to NGS, as NGS sequencing would need a good coverage of several reads on that MB for getting the same quality...

Also manufacturers claims of price and output doesn't necessarily hold true in real lab work.

[pmiguel]

SOLiD4hq? Why not the SOLiD5500XL? Actually I don't know what the "hq" suffix means. Our SOLiD4 was hauled away by Applied Biosystems when they installed our SOLiD 5500XL. I never saw the "hq".

As a rule of thumb: per megabase of raw sequence it works out that Sanger is 50-100x more expensive than 454 and that 454 is 50-100x more expensive than Illumina. Google "NGS Field Guide" for more detail.

There is something to what ulz_peter writes. Sanger sequence is probably worth at least twice as much as the equivalent amount of 454 sequence for many applications because of its (slightly) longer reads, possibly higher accuracy and very much lower indel error rate. That said, a single 454 run, taking 1 day, produces more sequence than a 3730XL run non-stop for a year. So, add to the costs of running that Sanger sequencer the cost of paying a technician 1 year of salary to produce that amount of sequence.

--
Phillip

[mbblack]

The data in the NHGRI plots is recent - http://www.genome.gov/sequencingcosts/ although it is not broken down by specific platform.

[colindaven]

Thats right, don't bother with the Solid 4hq since it doesn't exist!

It seems it was the projected successor to the Solid 4 before they redesigned and rebuilt the whole instrument.

The Solid 5500xl has now been out since spring/summer 2011.

[GenoMax]

It may be best to clearly state limits of what you want to use for calculating the cost.

Are you not going to account for the capital cost of the machine itself and/or the associated staff effort needed to run the machines. Are you calculating costs purely based on reagents/disposables needed?

Many academic institutions probably subsidize the capital costs directly or indirectly, so it would be difficult to come up with the "true" overall cost.

And then there is the cost of the associated informatics ... But if you are only looking at the cost of sequence generation then that would not need to be considered (though a part of the informatics costs could be considered essential for data generation).

[Chipper]

$33 for the SOLiD would be per Gb, not Mb, and most likely would only cover the sequencing chemistry. ePCR is not cheap, but I guess they forgot to mention it... HiSeq after the latest upgrade is ~$47/Gb including cluster generation.

[Graham Etherington]

Thanks for your replies so far.
To clarify, I'm only interested in just the library prep and chemical reagents cost of sequencing.
As to the point of sequence quality (i.e. Sanger being better 'quality' than NGS), I've covered information about the quality of sequences and the differences between the various techniques in the article.
Chipper - oops! yes, $33 per Gb. Thanks.
Cheers,
Graham

[strob]

could be of interest to read before starting a new experiment:

Genome Biology 2011, 12:125 doi:10.1186/gb-2011-12-8-125
Published: 25 August 2011

The real cost of sequencing: higher than you think!
Andrea Sboner1,2, Xinmeng J Mu1, Dov Greenbaum1,4,2,3,5, Raymond K Auerbach1 and Mark B Gerstein6,1,2*

Abstract
Advances in sequencing technology have led to a sharp decrease in the cost of 'data generation'. But is this sufficient to ensure cost-effective and efficient 'knowledge generation'?

[RubyTuesday]

Hi all,

Great question. I´m just getting into using a 454 Junior and am trying to budget out reagent costs for the coming months.

Google "Koumandou-FEBS-30Sept2011" and you´ll find a Sep 2011 powerpoint presentation that has some information on costs.

Let us know when you publish.

Cheers.