The news (thanks @danravicher) from the recent Illumina vs BGI_MGI that the invalidity of the ‘444 patent means BGI_MGI can enter the US market in 2022 was very interesting. Does it mean Illumina are worrying about new competition – I am not so sure. The cost of sequencing has been pretty stable for the last 6 years. And the last drop was a reduction in prices from Illumina rather than an increase in yields (this was the first time I am aware of that Illumina dropped prices).
There are many new entrants to the NGS space other than BGI_MGI; Apoton Bio, Element Biosciences, Genapsys, Singular Genomics and “They-Who-Shall-Not-Be-Named”, which should be great news for users. But two questions are uppermost in my mind:
-
Does Illumina need to compete?
-
Can Illumina compete?
NB: Before answering those questions I’ll quickly point out that I am NOT going to discuss Oxford Nanopore as a competitor – I’d suggest you take a look at Albert Vilella’s recent Twitter thread on ILNM vs ONT. And if you want to know about all the companies out there offering NGS jump over to Sandiegomics (nice new site Shawn).
Firstly, does Illumina need to compete?
Certainly, right now they don’t. No-one takes significant market share until BGI_MGI get into the US. And the rest of the start-ups have got at least 12-24 months before we really know if they can compete or not.
But even if one of these companies can compete it does not mean users will switch! There is a huge investment into Illumina tech in the library prep know-how and innovation (arguably most of which was driven by users and not Illumina) and the downstream primary and secondary processing. If someone offers a $300 genome it won’t be worth many people switching until they KNOW Illumina will not compete because it probably means switching at least some of, and maybe their entire, pipeline. And even then, if NewCo can’t offer the broad suite of applications then it may be better for them to ship out large e.g. WGS or exome projects.
Secondly, can Illumina compete?
I have speculated before that unless @illumina can do a big lift on SBS/Nova yields then they’ll be more reliant on lower prices than higher yields. Financially I have no doubt tat Illumina can compete if they choose to as their margins on sequencing reagents have been reported as 70-90% (although I am not sure anyone outside Illumina genuinely knows). If they can drop price then the questions would be, how low will they go and how will the stock market react to a reduction in margins?
Technically I don’t know if Illumina can squeeze another leap in yields on the Solexa tech without buying in technology that fundamentally improves on the patterned flowcells or imaging. And I previously speculated that a move by @illumina to mostly @GrailBio revenue allows them to bow out of selling sequencing/sequencers entirely and/or license SBS to @ElemBio et al. Could someone else make the boxes and reagents under license?
What happens at $100 or less:
Several companies have talked about $100 genomes, including Illumina’s CEO Francis deSouza. This is a truly disruptive price. For me, in a cancer genomics setting, it probably means we stick at $1000 per genome but increase coverage from 30x to 300x and ditch WES for tissue, whilst moving to WES for discovery ctDNA efforts. Just imagine the impact of a C2i Genomics or Foresight Diagnostics MRD approach, which would be cost effective and massively simplified over personalised MRD methods. For labs with small budgets it means experiments become affordable. For the big genome centres and all the funders it means saving a ton of cash on reagent spend.
But for a company like Illumina, and I’d probably argue any of the other NewCo’s, it creates a tough market where you’ve got to sell a lot of $100 genomes to make a decent profit. Of course $100 genomes could mean ten-fold cheaper single-cell, spatialomics, HiC or similar. So fingers crossed we’ll see the $100 before 2025!
In summary, I think user inertia is a huge challenge for NewCo’s to overcome – but Illumina have done some groundwork to give their users reasons to switch.
Great analysis!
How do you think ONT plays into this? Their PromethION is already cheaper with lower price/Gb and newer models with more flowcells would also increase output of a single run.
Great analysis!
How do you think ONT plays into this? Their PromethION is already cheaper with lower price/Gb and newer models with more flowcells would also increase output of a single run.
I intentionally left ONT out of this post. But that is NOT because I don’t think they can offer competition to Illumina. I’d focused on the new wave of ILMN “me-toos” that offer something more directly comparable. Check out What @albertVilella said (link in the post).