Illumina vs Pacific Biosciences : choisir l’entreprise de séquençage génomique de nouvelle génération

Séquençage génomique de nouvelle génération

Medicine is on the verge of a revolution driven by an exponentially growing knowledge of the human genome.

Un génome est la séquence génétique complète d’une personne. Le premier génome humain a été séquencé en 2003 après 13 ans de travail acharné. Le projet avait coûté 3 milliards de dollars. En 2007, il ne coûtait « seulement » 1 million de dollars pour séquencer un génome complet. En 2014, seulement 1 000 $. Aujourd’hui, le coût est d’environ 600 $.

Vous avez peut‑être entendu parler de la loi de Moore, qui a fait des ordinateurs une partie de notre quotidien. Cependant, la baisse des coûts du séquençage a été encore plus spectaculaire.

Ce progrès a été rendu possible par le NGS : Next Generation Sequencing. Il utilise les avancées des puces en silicium, mais au lieu de lire du code binaire, il lit le code génétique. Et plus la fabrication des puces en silicium progresse, plus le NGS devient efficace.

This dramatic decline in costs had 2 consequences:

• Un flot de données, avec de plus en plus de génomes séquencés, couvrant toutes les ascendants et pathologies possibles. Des outils d’IA et de big data commencent maintenant à les exploiter.
• La possibilité d’utiliser le génomique comme outil de diagnostic chez les patients.

Source: Illumina

Avoir la carte génétique complète d’une personne facilite grandement la recherche de ce qui fonctionne ou non dans son corps. Par exemple, nous pourrions comparer le génome de cellules saines à celui de cellules cancéreuses.

Les chercheurs pourront également réaliser de grandes analyses et identifier ce qui rend une personne vulnérable à la maladie d’Alzheimer alors que d’autres ne le sont pas.

And this is not all. The same technology that was used to do the sequencing of the genome’s DNA can also be used to create whole new scientific fields:

• Métagénomique: séquençage du génome des bactéries qui vivent dans nos intestins et ont un impact énorme sur notre santé.
• Transcriptomes & Épigénétique: si le génome est le plan, les transcriptomes sont la façon dont le plan est utilisé en pratique et en temps réel. Cela sera probablement très pertinent pour résoudre de nombreuses maladies incurables et problèmes métaboliques.
• Agrigénomique: utilisation du génomique pour améliorer les rendements agricoles et la productivité des animaux d’élevage.
• Études écologiques: évaluation précise de la santé d’un écosystème et de sa diversité génétique.
• Biologie synthétique: création de nouveaux gènes, traits ou organismes entiers avec un objectif précis: par exemple, des bactéries qui mangent le plastique ou des algues productrices de biocarburants.

Historique des entreprises

As I said, NGS relies on modified silicon chips designed to read DNA in an automatized process. Only 2 companies are dominating the field: Illumina and Pacific Bioscience of California (Which I will refer to as PacBio for brevity’s sake).

Ilumina a été fondée en 1999, et PacBio en 2004.

The first mover advantage has for a long time strongly favored Illumina. Today, Illumina’s market cap is $34B while PacBio’s is $2.5B. Illumina currently has 21,000 machines working worldwide, while PacBio has more than 1,000.

Overall, Illumina is the dominant company in the sector, and it tried to become a monopoly. In 2019, Illumina tried to acquire PacBio for $1.2B, which would have taken out its only serious competitor.

The deal was blocked by the Federal Trade Commission over concerns over maintaining competition. Considering the current valuation of PacBio, it was also a good thing for the company’s shareholders.

Both companies operate a business model similar to the one of mainframe computers in the 1980s. Sequencing machines cost $1M or more and are complex to operate, requiring significant training and experience. As a result, their main customers are very large hospitals, universities, and large pharmaceutical companies.

We are now entering the genomic equivalent of the 1990s for the computer industry. Easier to operate and cheaper sequencers are opening new markets in diagnostic, smaller research institutes, and all hospitals.

The more costs decrease, the more we will see genomic entering into smaller testing facilities. This is very similar to the way PCR machines have evolved, going from research tools to mass utilization, like during the pandemic.

It is expected that by 2027, oncology will quadruple its use of NGS for diagnostics, going from 2% to 8%. It should also double for genetic testing and increase by half for reproductive health and research.

Even then, this would give NGS a market penetration of just 14% by 2027, leaving plenty of room for more market share growth compared to other technologies. This drives the growth for the NGS market at 18% per year, doubling by 2026-2027.

Source: Illumina

Source: PacBio

Technologie

The commercial dominance of Illumina does not mean its products are superior to PacBio. They have similar accuracy and similar costs. Both companies are also releasing groundbreaking new sequencing machines.

Source: Illumina

PacBio

PacBio has recently acquired Omniome, giving it an advantage in accuracy for reading very long or very short genetic sequences. Its Onso system will be specialized on short sequences, while its Revio HiFi System will be able to read long sequences and work 24/7, making it able to read 1,300 full human genomes per year (roughly 4/day).

Pacbio expects long sequences to grow much quicker than the broader NGS market and claims its systems are 10x more accurate than the competition.

Source: PacBio

I would add that PacBio’s marketing strategy on this product is rather aggressive, directly attacking Illumina’s equivalent products, as you can see in this article.

Interestingly, it could also bring down the total cost of genome sequencing. Without getting too technical, short sequences need to be reassembled in a comprehensive whole. So a $500 short read genome can quickly need extra analyses costing an extra $1000 to be actually useful as a diagnostic tool.

So a $950 Revio long sequence analysis, being more accurate, avoids this step and comes out cheaper for diagnostic application.

Illumina

Illumina is launching the NovaSeq X, expected to be sold out for the whole of 2023. The goal of the NovaSeq X series is to create a massive amount of new data.

This comes in combination with the Illumina systems running on an open platform and standardized data format, allowing for better comparison between scientific studies and more collaboration between researchers.

About acquisitions, Illumina has been trying to buy back a spin-off from its own company, called Grail. Grail can detect cancer very early just from a blood sample (liquid biopsy). This could a true medical revolution. The company was spun off in 2016 and has since raised $2B. Illumina is ready to pay 9.7B to buy it all back.

The problem is that the deal was blocked by the European Commission to avoid Illumina having a monopole on liquid biopsy. The American FTC also had objections about it. Illumina ignored it all and acquired it anyway.

Now, Illumina had to put aside a $609M for possible fine by the EU. And might be forced by regulators to spin off Grail again at significant losses. This reflects badly on Illumina’s management, as they should have, first, never spun off Grail and, second, managed better the acquisition and regulatory risk.

Strategic difference between the two GNS leaders

On the big picture, it seems PacBio is at last developing a distinct strategy from being just an Illumina follower, focusing on high-reliability data compared to the more large-scale data collection of Illumina. This is also the result of the market maturing enough for more sophisticated users to emerge, as well as niche applications like long sequencing being now large enough to justify dedicated machinery.

To keep with the analogy to the tech industry, if Illumina is the giant, all-purpose  Microsoft of genomics, PacBio is turning into a specialized, high-performance Apple of genomics.

Comparaisons commerciales

Situation concurrentielle

PacBio focus on high-precision long-sequence analysis should give it an edge in diagnostics for the next few years. PacBio management believes this could boost the company’s market penetration in human genomics from 5% to 10% and in oncology (cancer) from 1% to 5%.

Those targets might be a little aggressive, but I do believe that PacBio has, for now, an edge on clinical analyses benefitting from long sequencing methods.

Illumina’s advantage seems to be more on costs. The NovaSeq X system is expected to bring full genome analysis to $200. This level has long been forecasted as the price needed for clinicians to start doing routine genome sequencing.

This is also a price low enough to justify some countries doing genome sequencing for their whole population, maybe starting with every newborn. Overall, this should boost the market for NGS massively.

I personally expect the demand growth might surprise everybody, including the management of Illumina and PacBio. Quite in the same way that the rise of PC caught everybody by surprise in the early 1990s. I would not be surprised that in 10 years, we routinely check out genome data to help design customized diets, optimize drug dosage, or calculate our vulnerability to specific infectious diseases or cancer.

Évaluation

Illumina’s stock price is more than 50% down since its peak in juillet 2021 and has gone sideways for the last 6 months. Even then, it is still rather expensive at a price-to-sales ratio of 7.3.

It has been growing revenues at 14.1% yearly over the last 10 years and is still growing aggressively.

(ILMN )

The question about the stock is more about immediate profitability. The cost of Grail acquisition and possible EU fine of up to 10% of the company’s turnover + forcing Grail spin-off loom over the stock.

Still, the company controls the large majority of the NGS market, which is poised to double in 4-5 years. It also has by far the largest budget for R&D and is on the verge of dividing the cost per genome by another 3x. So, as long as the company continues to perform in terms of price/genome, I expect it to stay the standard in most genomic research and for general medical purposes.

PacBio had an even more eventful stock price action in the last 2 years.

In juin 2020, it was at $3.73/share.  In juin 2021, it was at $34.97, boosted by the overall surge in biotech and genetic testing triggered by the pandemic.  To then fall back to $4.37 in juillet 2022 and recover to today’s $11.21.

(PACB )

The company’s long-term growth profile is “only” 9.9% growth of revenues in the last 5 years.

By all metrics, PacBio seems more richly valued than Illumina: its sales-to-price ratio is a very elevated 18.3; since 2017, it had only one year with profit in 2020, with the other registering losses in free ash flow and net income.

Thanks to a large $900M investment by Softbank in 2021, the company is not at any risk of running out of cash, even if not currently profitable.

Quel choisir ?

From a technical standpoint, both companies are great. They have very advanced technology and no real competitors except for each other.

Some startups are obviously trying to break in, like, for example, Ultima Genome, but this will be a tough battle to enter a very conservative market, as the struggle of PacBio to grow out of Illumina’s shadow illustrates.

I have used the analogy to tech companies in the 1980s-1990s. I think Illumina is similar to Intel or Microsoft at that period. They have great core technology and the advantage of a massive user base, de facto setting the standards for the industry.

The complaints of regulatory authorities remind me of how Microsoft was accused of being too large and too powerful at the time. How the situation around the Grail acquisition gets sorted out might be a catalyst for lower price or it bouncing back.

Decreasing costs and new use cases drive the market growth, which Illumina can then capture in the majority. Due to its dominant position, it is also the prime beneficiary of a larger market on which to spread its fixed costs of manufacturing and R&D.

In that context, investors in Illumina are likely to get good returns in the long run, benefitting from the persisting company’s dominant position. Risk-averse investors might prefer to wait on definitive news about the Grail acquisition future before putting their money at risk.

In contrast, PacBio is a riskier bet. It has always struggled to be profitable and could have been getting in financial trouble if not for the injection of cash in very favorable conditions by SoftBank at the height of the pandemic biotech rush.

Despite this troubled past, PacBio has been able to not only keep up with Illumina but even become technically superior to it in some aspects. Overall, I consider this an achievement in itself, as PacBio runs on an R&D budget vastly smaller than Illumina’s.

Because of this, I expect PacBio’s strategy to aggressively expand in oncology and human health to be rather successful at first.

Truly, the analogy holds, and PacBio seems like the Apple of the 1990s. Remember, Apple was a company struggling in 1995-1997, on the verge of bankruptcy, and saved at the last minute with an injection of cash.

It became the giant it is today through lean innovation and the development of a technologically superior product, appealing to a niche of professionals looking for optimal performance more than cost savings.

So, I believe that PacBio is a higher risk, higher reward type of investment than Illumina. It might fail in its ambition to push the industry to adopt longer sequencing as a method for diagnostics. Or it might succeed and grow quicker than the already very quickly growing NGS sector.

In any case, this is an investment more fitting for seasoned investors, as it might be highly volatile. A good technical understanding of the subtlety of sequencing technology can help as well. This is important because PacBio’s long-term success will be tied to its ability to stay ahead of Illumina’s base sequencer accuracy.