Tuesday, November 24, 2015

Well, that was brief!

BGI news today is that they are jettisoning the Revolocity large sequencing system, announced all the way back in June.  Along with the product abandonment, 40% of the ex-Complete Genomics group in the Bay Area is being laid off, with remaining staff focusing on the desktop BGISEQ-500 sequencer.

Revolocity was BGI's attempt to compete with the Illumina X platform, aiming for "population-scale" human genome sequencing. An open question, at least to pundits such as myself, is how much market is left for such huge boxes, with their accompanying huge run rate.  BGI's "strategic shift" suggests that BGI was no longer convinced that Revolocity made economic sense, particularly since much of the platform (at least at the level of chemistry) is shared with the BGISEQ-500.

The market for desktop sequencers remains hot; earlier this month QIAGEN finally launched their GeneReader platform.  QIAGEN had long frustrated genomics watchers by not forecasting any conventional performance metrics for their box, such as read length or number of reads.  This pattern wasn't broken: QIAGEN still hasn't described them.  Released at a molecular pathology meeting with a suite of QIAGEN-developed targeted assays, this is not a box aimed at genomics geeks for developing their own methods, but instead a closed ecosystem which QIAGEN will control from head-to-toe.

These devices join the Ion S5 and Illumina's MiSeq and NextSeq.  Illumina, as usual, dominates the market -- and the stability of Illumina stands in stark contrast to BGI's hasty exit from the high end.  Certainly a point to consider before purchasing a BGISEQ, though since it apparently is still aiming only at the Chinese market that isn't an issue for me (nor do I have budget).  MiSeq has been a wonderful platform for about five years now, but perhaps with the added competition Illumina will spring a major revamp in the next year or so.  Certainly Illumina could use to tune up the 2x300 MiSeq kits, which have sometimes had quality issues but this spring and summer became essentially unusable.  This was evidenced by the 2nd read's quality and accuracy cratering early in the run and the frequency of G & C diverging widely on samples.  NextSeq has seen significant popularity, so much so that the high output kits apparently have a tendency to become significantly back-ordered.

So the midsize short read market appears to be the emerging battleground for sequencers, with BIORAD/GnuBio's amplicon sequencer, Roche/Genia's nanopore device, and Genapsys potentially waiting in the wings, though the Roche/Genia and Genapsys devices have about as much in the way of announced specs as QIAGEN.  Is there enough latent demand to feed all these boxes, or will some of these also bail ala BGI -- or never even launch.  Stay tuned!

(correction from original: BGI made no formal announcement -- GenomeWeb broke the news)
(Correction 2: BGI/CG did not disclose the number of layoffs, not the 40% -- I need to fact check better!)


4 comments:

Anonymous said...

absolutely the right decision. it was obviously a pile. it probably only got to where it did through internal politics and short-term survival instincts. it's always the way.

Dale Yuzuki said...

Good morning Keith, one important distinction between the Revolocity and the BGISEQ-500 is the read-length format, indicating two different chemistries.

cPAL (ligation) on the Revolocity was updated to a mate-pair, 2x28 format with an average insert length of 375, although Rade Drmanac told me that the mates can be as far as 1kb apart.

cPAS (sequencing) on the BGISEQ-500 was 1x50, 2x50, 1x100 and 2x100 formats, as regular paired-end sequencing (not clear on the insert sizes).

Will be interested to see how all this plays out - the more competition the better, as prices will only come down as a result of it.

Marcel R. said...

Hi Keith,

Interesting read - we use a MiSeq with the 2x300bp chemistry and I was wondering if you could give more details to your observation with the MiSeq as I would like to check this as well for our application.
Thanks in advance!

Best,
Marcel

Anonymous said...


To be honest, genetics at current stage is mostly for research. Ordinary people won't give a damn to genetics. It's very close to useless to the mass. And for that reason, genetics companies usually cannot get stable revenue. They usually bankrupt at the end.