Thursday, January 14, 2021

JP Morgan: Illumina

Illumina presented at J.P. Morgan on Monday, reminding us that they aren't just a sequencing instrument company but an interlocking set of businesses focused on genomics. CEO Francis deSouza spent much of his time discussing the Grail acquisition and some of the other ways in which Illumina is pushing rapidly to become an essential part of clinical medicine, but there was one slide on future improvements to sequencing technology and a few on the lineup of existing sequencers.  Reminder: I'm working off public sources, as during the day we work closely with Illumina and they even sunk some serious cash into my employer last May.

First off, the "I'm not dead yet" reminder that Illumina still has a microarray business, growing at 5% annually.  The mere existence of this business is a humbling reminder that sequencing snobs like myself expected that entire technology to melt away about a decade ago, yet still it plugs on.

deSouza had some bold predictions for Grail, which has developed a cancer early detection screening assay called Galleri.  In particular, he believes that Galleri will have as big a positive impact on patient outcomes as all current therapeutic options.  That's quite a claim!  Grail has clinical trials ongoing to try to start proving that case.  The question as always will be if early detection really means higher treatment success and lower cancer death rates, or conversely simply longer times spent living under the shadow of cancer treatment.  I'm at that age where I lose a contemporary (broadly speaking) to cancer about once a year, so it is a topic of significant personal interest.  Last year it was a sysadmin a decade and a half my senior whom I respected deeply, taught me much -- and locked horns with periodically. 

deSouza touted growth in many areas, both through Illumina's own efforts nd numerous partners. Non-invasive prenatal testing (NIPT), oncology and rare diseases all were highlighted with impressive numbers.  Over 1 million pregnancies a year screened by NIPT and numerous partnerships for companion diagnostics in oncology.  He mentioned as one example of the effect of lowered sequencing cost -- $600 human genome is their current claim -- is that Shriner's Hospital for Children has both switched from exomes to genomes for rare genetic disorder diagnosis but also expanded this to all 22 hospitals in their network.  Denmark has signed up with Illumina for population-scale sequencing -- and Illumina is now selling "Illumina Connected Analytics" to assist such endeavours with population-scale analytics.

On the instrumentation front, Illumina highlighted placing over 2000 instruments and gaining more than 700 new customers.  NovaSeq is the first blockbuster instrument, generating over $1 billion revenue.  As noted with Shriners, Illumina believes that ongoing improvement in price/performance ratio, such as their version 1.5 flowcells released last year, are creating new business (aka "demand elasticity").  deSouza noted that they have sold a record number of their top-of-line NovaSeq S4 flowcell.  There's still 320 HiSeq customers which Illumina is trying to woo onto NovaSeqs.

It is interesting to note that not much time was spent on the entry level systems -- iSeq, MiniSeq and MiSeq -- other than to say that these bring in new customers.  Oh,  and that the new NovaSeq 2000 P1 flowcell, rated at 20-40Gb and shipping in the second half of this year, will entice MiSeq and NovaSeq 550 shops to upgrade to NextSeq 2000.

So is Illumina getting set to cede the low output short read market to the likes of Genapsys or waiting-in-the-wings Omniome and Element Biosciences?  First, just saying "<20Gb" is low output feels a bit absurd, given that Genbank didn't hit 3Gb until I had been in this field for almost a decade.  Genapsys -- presenting today simultaneously with PacBio (and a meeting I can't miss!) -- wants to get into that P1 space if they can launch their 144 million cluster flowcell, and their box is a tiny fraction of the NextSeq 2000 to buy.  Will Illumina see these upstarts as still seeding interest in Illumina's big boxes, as pipsqueaks unable to compete with Illumina's marketing muscle, or threats to be sniped at with IP lawsuits and other tactics?  Will Illumina reinvest in entry level instruments to cover the sub-20 gigabase market?  Was iSeq a one-and-done technology with no future?

On the technology front deSouza did dangle some interesting numbers.  He showed micrographs of the current NovaSeq flowcell nanowell geometry and a 5-fold denser packing.  Could this be handled on existing instruments or will it require upgrades?  That wasn't covered.  Since this was a financial crowd, he also mentioned manufacturing improvements that allow more flowcells per wafer -- and a 90% reduction in production costs.  Obviously that won't lead to a 90% reduction in flowcell prices; Illumina will be happy to pocket much of that and perhaps use a bit for pricing flexibility.  

He also mentioned demonstration of two-fold improvements in read length and cycle times and three-fold in accuracy, pinning this on "new dyes and blocking chemistries".  So will there be 500 basepair paired end, or was this "just" a hint of 300 basepair paired end on NovaSeq or NextSeq 2000 class instruments?  All left ambiguous.  But do start tuning your pipelines and databases for a future "S20" flowcell delivering 50 billion fragments and 30 terabases per run, as that is a very real possibility for next year's JPM.   If I had to guess, it will require a next generation NovaSeq sporting optics technology developed for NextSeq 2000.

What flies are there in the ointment?  deSouza did spend a lot of time discussing pandemic effects on their finances and that it will take a while for revenues to return to their previous trajectories post-pandemic.  He didn't discuss the new IP countersuit from BGI or the low end threats.  There's also the constant challenge of talent -- resident genius Mostafa Ronaghi announced his retirement in another forum.  Mostafa has an astounding intellect in sequencing space (certainly elsewhere as well; I've only talked sequencing with him); we had some discussions with him at Warp Drive and anything we asked about he had some preliminary approach right at his fingertips.  There are Illumina alums all over the genomics industry; that means a constant battle to backfill in addition to supporting growth.  Illumina also both serves some large partners in the clinical space and competes with them; it would hardly be shocking if some of those companies were itching to invest in an alternative.  deSouza also basically whistled past the long read space, saying that their superior analytics were enabling short read solutions for what were previously thought to be long read exclusive problems.   Finally, we know that Illumina won't find it easy to acquire any other sequencing technology companies, given the gimlet eye they got from regulators on the PacBio attempt.

1 comment:

new said...

Interesting as always!

Unfortunate that there's no scale on the SEM images... The last SEM images I saw showed 500nm wells. So 5x denser would imply 200nm or smaller? I guess they've deployed their super-resolution approach across most platforms now...

Illumina seem prefer selling high-end instruments. Their margins are likely better, and field support requirements lower.

Interesting that they focused on clinical medicine, and Grail/NIPT. I think they've realized that their IP monopoly in sequencing is up and they need to expand downstream if they want to maintain a hold on sequencing in clinical apps.