[xyzzy123]

I had guessed that January's drop was connected to Illumina's announcement in the same month, so was surprised that it reversed in April.

[AllSeq]

I had guessed that January's drop was connected to Illumina's announcement in the same month, so was surprised that it reversed in April.

My guess (and it's really just a guess) is that they are ignoring the X Ten in figuring the prices. Illumina has been raising prices on their sequencing kits every April 1st, so that little bump makes sense. However, I'm not sure that they're taking into account the release of the v4 chemistry for the HiSeq 2500. That reduced the list price for 1Gb of sequence from $43 down to $29. (For comparison, it's just $7/Gb for the HiSeq X.)

[xyzzy123]

https://www.genome.gov/sequencingcosts/ is no longer being updated. It looks like they hit the end of the graph and don't want to extend it. In any case, since the price is now dropping much more _slowly_ than Moore's Law, in another 10 years or so at this rate Moore's Law will catch up and the graph would look kind of silly.

[xyzzy123]

NOW there's finally a sizable drop.

https://www.genome.gov/sequencingcosts/

The last two entries in the .xlsx file show a cost per genome of $4,211 on Apr. 15 and $1,363 on Jul. 15. I guess they decided to take the X Ten into account.

[Brian Bushnell]

Due to all the artificial limitations on the X10, I don't consider it an accurate reflection of real sequencing costs.

[westerman]

Due to all the artificial limitations on the X10, I don't consider it an accurate reflection of real sequencing costs.

Indeed. Those costs are misleading. There should be all sorts of asterisks detailing the assumptions. Might as well throw in the higher per-base cost of using PacBio to complete a genome.

I've always hated that graph because of the misapplication of "Moore's Law". Stupid. But I have seen it in so many presentations that I just grit my teeth.