$256 Million Worth of Spit

Managing Editor’s Note: We’re not far off from Jeff’s Pitch Room…

In just a few days, he will reveal the details of the tiny company that has already drawn investments from the founders of ChatGPT and even the CIA, thanks to its AI “super chip.”

This chip is 50 times more powerful than NVIDIA’s H100 and could shake up the whole industry once it goes mainstream…

So, make sure you go here to automatically add your name to join him this Wednesday, May 21, at 10 a.m. ET.

And the winning bidder is…

Regeneron Pharmaceuticals.

In a court-run auction for the assets of bankrupt 23andMe (MEHCQ), the biotech giant prevailed with a winning bid of $256 million.

It’s actually a remarkable price, considering – prior to the announcement – the enterprise value of 23andMe, now trading on the Pink Sheets, was just $5 million.

As a reminder, 23andMe has long operated as a consumer genetics company. But it had much larger aspirations for the business… more on that in a bit.

Founded in 2006, it was very early in the movement to make genetic sequencing and analysis services available to the consumer market at an affordable price.

23andMe is still operating its consumer genetics services for both health and ancestry, a service which will continue to be operated by Regeneron after the acquisition is complete.

23andMe Service Offerings | Source: 23andMe

I was tracking the company as early as 2007 and actually used 23andMe to sequence my own genome in October 2011.

Jeff Spits in a Tube

Shown below is my confirmation e-mail from 23andMe.

Note the date and time stamp in the upper right corner – October 25, 2011.

(Click to expand)

I was just being curious back then, checking out – what was then – a company at the leading edge of genetic sequencing and analysis.

I didn’t learn much.

In fact, I would argue that I was paying to contribute to a growing genetics database that might become useful at some point in the future.

And I was OK with that – paying to contribute my DNA for science.

I’ve long believed that the key to personalized medicine is the combination of genetic sequencing and artificial intelligence.

A massive database of high-quality DNA sequencing can ultimately be used to understand what genetic mutations are responsible for disease and what genetic sequences result in a healthy human condition. And that information can be used to develop individualized genetic therapies optimized for an individual genome for both efficacy and safety.

And in the very near future, the cost to sequence a whole human genome will drop to $100 and eventually will be covered by health insurance.

In time, all children, with parental consent, will be sequenced when they are young so that physicians can gain a clear picture of the child’s health.

Understanding if there are any mutations that would knowingly lead to a disease could be the difference between life and death for a child. And at a very minimum, it can inform the family of environmental changes to manage the disease, or potential genetic therapies to cure the disease.

For example, there are already CRISPR-based genetic editing therapies approved by the FDA and European regulators for beta thalassemia and sickle cell anemia. One injection and the patient is cured. Some would call that a miracle, but it’s just science. Removing the mutation responsible for the disease and replacing it with a healthy genetic sequence cures the patient from the disease.

The challenge of realizing that future was, of course, where might one get all the data?

After all, sequencing whole human genomes isn’t like appropriating data via the internet through surveilling, which most of us don’t even realize is happening – it’s just software running in the background.

Collecting whole human genomes first requires a vial of spit or a blood draw and then a subsequent sequencing on an Illumina (ILMN) gene-sequencing machine. (Illumina controls over 80% of the next-generation sequencing market.) And this needs to be done at massive scale.

And this is what most people, allured by the prospect of discovering the origins of their ancestors, didn’t realize…

23andMe was just a front for collecting DNA to monetize the DNA through drug development.

The Elevator Down

23andMe went so far as to sign a deal with pharmaceutical giant GlaxoSmithKline (GSK) back in 2018.

The deal provided 23andMe $300 million in much-needed funding in exchange for equity ownership.

At that stage, 23andMe had already collected five million consumer genomes (more on this in a moment)…

And GSK was planning that this $300 million deal would “usher in a new era” of genetics-based research at GSK.

As it turns out, it was money poorly spent…

Stock Chart of 23andMe | Source: Bloomberg

GSK’s investment back in 2018 – at a $2.45 billion valuation – was looking positive for a couple of days, after 23andMe went public via a reverse merger with a SPAC and enjoyed a $6 billion valuation… momentarily.

But as we can see above, GSK rode the rollercoaster down with 23andMe, collapsing 99.75% from its all-time highs – straight into bankruptcy, which was declared this March.

The problem?

Well, from my perspective, there were two.

Where the Magic Isn’t

Perhaps the most ironic part about the whole business is that the 23andMe CEO, Anne Wojcicki, was married to Sergey Brin. Brin is the co-founder of Google.

And so naturally, 23andMe was leaning in hard and trying to collect as many DNA samples as possible, as quickly as it could.

Wojcicki was simply applying many similar principles to data collection and monetization as Google did, just in the life sciences space.

I can imagine the framing that was used in investor presentations made by Wojcicki when she was raising capital. I’m sure the analogs between the two companies were clearly spelled out…

But applying internet economics in life sciences is a fool’s errand. The return on investment is very slow, there’s a lot of friction, and the economics are very different.

23andMe was never able to achieve profitability, nor was it able to come anywhere near becoming cash-flow breakeven.

Hence, bankruptcy.

But the second problem was the quality of the data.

23andMe was collecting DNA, sequencing it, and sending back information with very limited value, and typically no actionable information.

In its collection, it only collected phenotypic data as self-reported by the consumers, not clinically collected phenotypic data.

Our phenotype (i.e., the observable and measurable characteristics of our human condition) is the most critical in understanding our human DNA.

This is where the magic happens. Understanding the connection between the genome and the phenotype is where we discover the specific genetic sequences that lead to both good and bad health. And having these two data sets linked is where artificial intelligence comes into play.

This is, in fact, what drove me to Human Longevity back in 2015. Its strategy was exactly that, and I put myself through their program back in 2020 and have been benefiting from that process since then.

That invaluable, combined database has since been spun out into a private company that I’m very bullish on, Simplify Genomics, which is empowered to achieve something far greater than 23andMe aspired to.

The bankruptcy of 23andMe is interesting for two reasons.

Ripe for Revolution

I believe that when we look back a few years in the future, we’ll see that 23andMe’s bankruptcy marked the bottom of the biotech winter.

23andMe generated more than $200 million at almost 53% gross margins in 2024. And it has invested more than $1 billion getting the business to this stage. There is a real business there, and 23andMe did achieve impressive scale, and it was picked up for a mere $256 million at auction.

Even without the high-quality phenotypic data, with artificial intelligence, there was still value that could be created. In a healthy biotech market, 23andMe would have had no trouble raising additional capital.

The other interesting point is that 23andMe, despite trading for just a $5 million valuation prior to the auction, could not come to an agreement to be acquired.

Late last year, 23andMe announced that it had entered into a process to look for strategic alternatives, which meant that it was accepting acquisition offers.

23andMe had no leverage at the time, as it was out of cash with no prospects for funding. If it did receive any offers, they were clearly too low to consider.

This is precisely the issue with the biotech industry today.

More than 100 biotech companies continue to trade at negative enterprise valuations, despite being worth eight, nine, or even ten figures more.

It literally took a bankruptcy to have price discovery of 23andMe.

And that price was $256 million, which is $251 million more than its pre-auction valuation.

I can’t overstate how indicative this deal is of where we are today in biotech…

A company that did, last year, $200 million at almost 53% gross margins…

Was trading at a $5 million valuation.

Regeneron (REGN) got a deal.

Back in 2013, it established its Regeneron Genetics Center, and since then has discovered more than two dozen novel drug targets using its own genetic database, comprised of about 2 million human genomes. Now, it has 15 million more.

This genetics-powered precision medicine and healthcare revolution is overdue. We have all the pieces now: low-cost sequencing, large databases, and generative artificial intelligence capable of reasoning.

Regeneron’s acquisition of 23andMe assets in bankruptcy will turn out to be a landmark deal and the beginning of something to look forward to.

Jeff