Ginkgo Bioworks Holdings, Inc. (DNA) Earnings Call Transcript & Summary

Perfect. Hi, everyone. This is Rachel Vatnsdal from the Life Science Tools and Diagnostics team here at JPMorgan. I'm joined by Jason Kelly from the Ginkgo team. And so today is going to be a 40-minute presentation. Jason is going to start off with some slides, followed by about 20 minutes of Q&A. [Operator Instructions] And for those of you in the room, please raise your hand. We have mic runners throughout. I just ask that please don't ask your question until we have the mic in your hand, so we can all hear you. And with that, Jason.

All right. Well, thanks. Jason Kelly, I'm one of the Co-Founders and CEO here at Ginkgo. I want to start by thanking JPMorgan for getting us all back together this year. It's actually -- I'm sure it was not easy to reboot this after a couple of years off, and it's really nice to see everyone. Okay. So I'm going to kind of cover 3 topics today. So the first is, 2022, from my perspective, was a banner year for Ginkgo and really synthetic biology technology in general. So I want to give you a few of the highlights of what happened over the last year. Second, we're at JPMorgan. So the kind of focus here is around therapeutics. Ginkgo operates in a wide range of markets, but I'm going to give you a deep dive on how some of our customers in the biopharma industry are using our platform for pharma discovery and manufacturing. And then finally, synthetic biology as a technology is becoming increasingly a national priority here in the U.S. And so I want to give a few comments on what I'm seeing there from a government standpoint. Okay. So this is our first year as a public company. So we did go public October '22, and -- I'm sorry, '21. And so from my standpoint, I think I'm quite proud of what the team has pulled off. So we nearly doubled the number of programs that we run at Ginkgo in terms of new programs added this year compared to last year. And so this is important for a few reasons. Every time we add a new program at Ginkgo, it adds 3 sources of value for us, right? So the first, we get some near-term revenues in the form of like service fees, right? So think of us doing contract research for one of our customers. The second is, all of these programs include some form of what we call downstream value share. So we're either getting a royalty or we're getting equity in a small company or a milestone payment upon success of the work we're doing for a customer. And then the third reason programs are critical at Ginkgo is, we get intellectual property and data. So the way we work out our IP arrangements with our customers is, Ginkgo can reuse the data and learning from the programs we do in areas outside of what the customer wants it for. And so I'll highlight some of those data assets as we go through the talk. So the fact that we've been able to double the number of new programs added is a critical metric for me internally at the company. The other thing I'm quite proud of is that we're targeting $460 million to $480 million in revenue this year. And today, we're able to say, we're reiterating that guidance. That's really exciting. If you look back to when we first gave our guidance for the year in terms of total revenue, that's about a 35% increase from the high end of the range we shared back in March of '22. So for example -- and again, a lot of work by the team going into delivering on those numbers, again, as our first year as a public company. And then finally, not lost to anyone in this room. It's a tough capital market for growth companies and for biotech growth in particular. We're ending the year with about $1.3 billion in cash, which allows us to continue to operate quite strategically in the market. All right. So like what happened in '22, I want to give a little bit of an insight for those of you who are new to Ginkgo in terms of just how we operate. So if you put on a pharma lens, I think an easy way to think about Ginkgo is, we operate as sort of a CRO, right? But in contrast to traditional CRO where you're going -- and basically, biopharma company will go and look to outsource, I would say, lab work they don't want to do, okay? So I don't want to do an animal study, I'm going to call Charles River. I don't want to do the synthetic chemistry project, right, I'm going to call WuXi. When folks call Ginkgo, they're looking to outsource lab work that they can't do internally, okay? They want to get access to a level of automation scale that they don't have internally. They want to get access to a set of data and machine learning models or genetic assets that they don't have internally. And you can see that in some of the comments from our partners, we announced is here Marcus Schindler posting on LinkedIn, the CSO at Novo Nordisk, got for our deal with them earlier this year that they're looking for external partners that bring new and complementary expertise. We're seeing this increasingly in biopharma and other spaces where folks are looking to get things that they don't have internally. Similarly, we did a deal with Bayer in the agricultural division last year where you see them outsourcing really their microbial biologics work to Ginkgo. And that's a large project, multiyear. Again, part of the same trend of outsourcing, not to do work you don't want to do, but to outsource to get access to technology you don't have in-house. So the other thing that happened in 2022, we completed 8 acquisitions. And one that I want to highlight, particularly critical, a company called Zymergen, based here in the Bay Area. And one of the technologies I'm showing you in this video is their automation platform technology. And at the start, what you saw were little carts that could basically be plugged into this magnetic lab track, where you see samples being moved, parking in front of these -- I'll put it again, parking in front of these rail locations and then an arm being able to pick up the sample off that track and move it on to a piece of equipment. So why is this important? Those carts can be swapped in and out, and that allows you to have a high throughput operation that's also flexible for us to change the type of equipment that we're plugging in. That ends up being very important given the range of microbiology activities you need to do in cell engineering. We have this little -- interested in stickers or hats or shirts with this No Pipettes logo on it. I'd be happy to give them to you. I come from this world, right? I did a PhD at MIT. It's 5 years of learning really how biology works, taking really smart people, making them into good biological engineers and then also teaching them how to do tedious manual labor at the lab bench, right? And that is an absurdity, right? So if you play out 5 to 10 years from now, you should expect total laboratory automation in cell engineering, right? The idea that a scientist should be picking up with pipette is absurd, okay? And so we are happy to be driving that transition. We think as you as a biopharma company are looking at your research infrastructure, you should not expect to see scientists at lab benches pipetting in the areas of synthetic biology and cell engineering in the future. We have some stats to back that up. So we measure the output of our facility. So here on the left-hand side, you can see a measure what we call a strain test. It's basically us. We built a strain, we've engineered it and we're doing some sort of assay on the performance of that strain. And so we're quite happy that we doubled the number of the strain tests that we did in the last year. I do want to highlight, if you look at the little blue dots, it's hard to see, unfortunately, but we've gone from like 50 to 100 strain tests a day back in '20, I guess that's 15, to now into the hundreds of thousands of strain tests per day. If you look at the meat of that line, but then you see these blue dots at the top that are 1 million, 5 million, 10 million strain tests in a day. And this is something we're going to touch on in some of our upcoming earnings call at Ginkgo, where we give a little more of a deep dive on this technology. But what we're seeing is a [ RAD ] testing on plates as well as pool testing, thanks to genomics and DNA barcoding, allowing us to tap much larger numbers in certain types of assays. And so I expect you'll see more of that in the future, and we'll be evolving these metrics as the technology changes. Alongside that, about a 20% drop in -- if you were to kind of look at those strain test costs across the company on a per strain test basis. And again, these are preliminary numbers, but I want to give you a sense. The other asset we have is what we call our code base. So remember, as we do these projects with customers, we're retaining our data and learning. We're also, thanks to these acquisitions this year, we acquired -- we've historically acquired Warp Drive Bio, but recently Radiant Genomics, Lodo Genomics as part of acquisition of part of Bayer's agricultural unit, we brought in a large strain collection. We now have, I believe, the largest proprietary gene collection in the world from microbes, over 1 billion gene sequences. This serves as an excellent repository to go looking for new types of enzymes, maybe new types of gene editors. I'll show you some data in a minute to tap into that database. And again, this is available to any customer on our platform. Importantly, when you go hunt for an interesting gene in that database, you then want to synthesize it, put it into the genome of a cell and test its performance. And so here's some data that I think is really cool. So on the y-axis here, you have enzyme type by EC number, so 1.1, 1.2, 1.3, going down from the top, okay? And then on the x-axis, you can see number of samples we've tested in that class of enzyme. And again, a bit hard to see at the bottom here, but those long rows go out about 5 million tests. Okay, so what that means is -- and the colors are different programs we've been doing for customers. So because we have the rights to reuse these programs, we've done all these programs for that first enzyme class at the top 5 million strain tasks, and then that is used to inform our machine learning and AI models. And I know if you talk to folks in biopharma, they'll tell you that they've been having all these meetings this week, they're hearing AI and ML so much, they want to stop hearing it. The important thing to keep in mind is, AI and ML algorithms at this point are a commodity. What is not a commodity is the data that trains those algorithms. So when you're hearing from a company about an ML or AI assets and they're running it on publicly available data, that is a end-to-end commodity object, and you should pay for it equivalently. If they're not showing you something like this, millions of data points that are going in to train those models, then you don't expect to get something that's different from the guy next to you, okay? And so this type of proprietary data is what comes out of those strain tests from the automation at Ginkgo that have doubled in the last year. So I'm really excited about this flywheel, and you'll see in a minute how it actually saves us time in the lab to have those models. The other thing I want to highlight is because we have these assets, customers are signing up. So I mentioned this earlier, but we went from 31 cell engineering programs back in 2021 that were added to the platform. And again, we're reiterating our guidance today of 55 to 60 programs and also highlight some of those are with existing customers. We love to expand with our existing customers and add new programs, but A number of those are new customers that are getting on the platform for the first time this year, both large customers like Merck and Novo, and also a number of smaller companies and both inside and outside of biopharma. So really happy to see new logos and that growth. Okay. All right. Now I want to dive in and give you a little bit more of a deep dive on biotherapeutics particularly for the sort of crowd here at JPMorgan. So the #1 point I want to make and is a big source of confusion about Ginkgo is, we do not develop our own therapeutics, okay? So there is not a drug pipeline I'm going to show you. I'm not going to show you preclinical results from our asset or this out or the other thing when I'm engaging with pharma -- biopharma companies. I'm not going to their BD team with an asset to hand them, I'm going to them with a technical capability. And there's upsides and downsides, right? Downside is people love the transaction assets in biopharma, and that's not part of my business. That's a downside. Upside, I can be in every modality, right? Because I'm not actually doing the downstream drug development in all these areas. I am a platform enabling all of my customers to be operating in all of these areas. And we basically can play anywhere where cell engineering plays. So anywhere we're designing the genome of a cell, making it do something new at scale is important. And it turns out that's basically everywhere, including in small molecules. And so you can see here we've listed some of our publicly announced partnerships, and this ranges from discovery efforts with Selecta on capsid design as well as manufacturing in AAVs, for example, with Biogen. On the biologics side, manufacturing with Novo. A number of discovery deals in the microbiome space, MRA space as well and so on. And so I'm going to now go and give you a couple deep dives in some of these areas. So you can see the type of data that biopharma companies would be seeing when they're deciding about working with Ginkgo, at least at a top level. We don't have a ton of time today, but I'll give you some of it. Okay. So in cell therapy, these are some of the areas we're working in, regulatory elements, novel CAR designs, IPSC engineering. And I'll just note, this is all driven -- we just announced the opening of a new, I think, 6,000 or 7,000 square foot facility in Boston with all dedicated robotic automation to mammalian cell engineering, okay? And so this is probably the fastest growing area of business for us right now, expect us to do a lot more here. But I want to do a specific deep dive on some data we shared at the SITC Conference, Immunotherapy Conference in Boston a couple of months ago in novel CAR designs. And so what you're looking at here are ICDs, so intracellular domains. And we have 2 sort of libraries we're building: one is a 2-domain library, and one is a 3-domain library. And you can see the 100 by 100 by 100 across the domains. So if you have 100 domains cross 100 domains, you get a 10,000 member library. 100 by 100 by 100, you get 1 million member library, okay? And so we've been -- so I'm going to show you the data from the 10,000 member library, but the 1 million libraries also in flight. So if you take a look here, what we do is, we start with this 10,000-member CAR library. We're using lenti to transduce primary T cells. And then what we're looking for here, like a specific demonstration we wanted to do for folks, was to look at the challenge of exhaustion. And so what we do is, we call culture with tumor cells. We allow this to expand. There's a 3-week period where we're looking to see how -- which designs persist. In another words, which ones resist exhaustion the most and keep expanding. And so you can see the data here, it's really quite nice. So we have sort of on each axis just a replicate of the same experiment. And the dark blue dots are these 55 clones, these 55 genetic designs where we see more than 30-fold enrichment compared to the rest of pool. And importantly, we do this both with a high affinity and low affinity extracellular domains, and we see sort of different results. And one of the things we want to do in the future, as you might imagine, is combine libraries of extracellular domains with intracellular domain. I think there's real cool opportunities there. And so I'm not going to show the data for it, but we did take these hits and then subject them to high-content arrayed tests and get a whole bunch of data about the sort of phenotypes of these cells that does give us some pretty interesting mechanism hypothesis on this. So we're really excited about it, and we're really excited about taking this to even bigger libraries. Some of the other things you should expect us to do coming up, we're going to try this under different cell culture conditions that are a little more tumor-like. I have time to talk about today, but we acquired a company called FGen, which does encapsulation in alginate beads that allows for looking at single cells. And so we're running that on the platform now. We're going to apply that to get more single cell data here. And also, as everyone asked us about here at JPMorgan, we'll be doing -- putting this into animal trials as well to get additional data. The last thing I will mention since I've got the attention of some biopharma folks is, we can take that library right now and put it behind your extracellular domain, right? So if you want us to try this, it really -- like this is all set up to run. Like in a matter of weeks, we could be trying these libraries behind -- in your CAR constructs and running a similar type of assay or other types of high-content assays at Ginkgo. So really excited about this. I think this is something we've seen a lot of interest in the market. The other area -- and these are mostly partnered programs, so I don't have as much data to show you here, but I just want to highlight we're working in this area is an AAV and gene therapy generally. Regulatory elements, as I mentioned, we're doing some capsid work with Selecta and others. ITR engineering, also working on the payloads and then also working on the challenge of manufacturing in AAV. We see opportunities in all of those areas for our platform and do have some partnerships there. This is one I think is pretty interesting, and I want to draw on what I mentioned earlier, which is we have this huge microbial genome collection. And so we did -- we've done this as a computational analysis, and there was a paper on it, get the name right, Makarova, in 2020, who did a similar analysis looking at NCBI and looked for computationally CRISPR-like systems based on what was out in the public data and found 6,700. We've done a similar analysis now at Ginkgo, where we find about 93,000 in our databases, and then you can break it up by type. So type 1, 2, 3, 4, 5, and you can see the enrichment relative to what's out there in the public databases 19-fold, 2.5-fold, 12-fold. So if you were out there, you were a gene editing company. I want to highlight, this is available to you, right? Don't feel like you're limited to just what you have access to. We're readily available. We operate with the CRO business model. You can access this on a service basis, and you can access it from an information content standpoint. But then importantly, we can find you a bunch of hits and plug them into all our automation and very quickly get your results on which ones actually work well or have functionality in the lab. And so -- and our model at Ginkgo is a small piece of a lot of pies. So if you think about kind of the structure of our deals, you should expect it to look different than if you were talking to a single asset pharma company, right? If I'm a small biotech and I've got my one asset, boy, do I want the world for it? That's not my model at Ginkgo. My model at Ginkgo is to be a utility. We want to be looking Amazon Web Services. We want to be working with every biopharma company on every project in all modalities. And if I'm going to do that, I don't actually need to take that much downstream value on a per program basis. So again, if you're in the editing space, I encourage you to reach out. This is fun. So we announced the launch of Ginkgo Enzyme Services. It's our first sort of like turnkey services offering. It's a little faster for us to do the deals. The deal term's a little more standardized. We announced actually a deal in this space with Merck recently, $133 million deal, to work on improvement of biocatalysts. In other words, enzymes, okay? And the request is often pretty similar. People want to see an increase in activity of the enzyme on the particular reaction that's catalyzing. So this one had a low synthesis to hydrolysis ratio, for example. And then they want to see less off-target, right? So more specificity in terms of what the enzyme is making. And so what's cool here is, we generated and screened just over 1,000 candidate enzymes. That's actually on the lower end of the numbers for Ginkgo. And one of the reasons is, these were not chosen randomly. These came out of structural and ML models that were informed by that EC data I showed you at the beginning of the talk, right? So because we have all this data to feed into our algorithms about how these enzymes perform, it allows us to actually do less lab work, which improves my economics and allows me to offer even better deals to our customers, right? So there's a nice feedback loop where the more we learn actually the less we need on a per project basis to do in the automation, it means we can get more automation to generate even more data in other areas. But we had great results here, 6.5-fold improvement in activity relative to what the customer's current best enzyme was and a 5x improvement in the specificity. One last point to mention on the enzyme side. There are also a number of companies developing therapeutics where sort of the enzyme is the point of the spear, right? It is really the drug. And so we have a project with Synlogic, which is working on, in this case, HCU is the disease. And what's really good about this is, the first Ginkgo-engineered organism that's gone into clinical trials. And so really excited about this. But in this case, it's an enzyme being delivered by a probiotic microbe for treatment of a metabolic disorder. But if you are in the business of trying to develop an enzyme as a biologic or as part of a microbiome therapeutic, again, I would say, our platform is very likely to be relevant to you, and we'd love to talk to you. Okay. Last thing I want to mention as we done this sort of pharma section is, the work we've been doing in circular RNA and RNA therapeutics and vaccines. So we acquired a company called Circularis recently. We're really excited about circular RNA. In the middle there, I just have sort of like some of the known challenges around RNA therapeutics, stability, durability, challenge of immunogenicity, length and size, expression level and specificity. We see a lot of opportunity to work on these problems through circulizing the RNA that is delivered to the patient. How do you circulize? There's actually a few different ways to do it, right? And so one of the things we're approaching here at Ginkgo is not to just bet on one thing. We want to kind of bring in a few different options and try them out in combination. So I'll talk about some of those methods, and then how do you dial in the expression strength, how do you get the expression of the thing you want, code on usage structure, modified nucleotides, translation elements. We can work on all those. And then we're actually -- obviously, we do a lot of work in manufacturing so we think we can be pretty immediately helpful there as well. So to just show you a little bit of data. This is on the lower left part of this triangle, the circularize. So using 2 different methods of circularization, catalytic introns, hairpin ribozymes. What you see here on the red dotted line is sort of your traditional capped modified poly-A mRNA. So like you take that Moderna shot, that's what you're seeing. And you see those declines on a 5-day experiment. The green lines are different -- some of our different circular RNA constructs using both of these 2 approaches. And again, this is early is not in animals and so on, but we like this data. You expect us to do more here, and we're getting a lot of interest from folks in the space on this. The top of the triangle was around expression. I just want to highlight, we've designed and tested a 500-member library of internal ribosome entry sites, right? So if you have a circular RNA, the RNA ribosome has to show up somewhere. So you use these IRES to bind and get started. We tried to a library there. We get different expression levels. Importantly, it also helps for this sort of specificity challenge. So doesn't express in what cell types, and you can see some of our data there, 2 different cell lines where we get preferred expression and then some where it's expressed in both. And so we can do this in times many more cell types. We can do this with many more IRSs. This we think is really exciting. Okay. So it was my quick dive on what we're doing in pharma as much as we could get through today. And I'm just going to end real quickly on where I see synthetic biology as a technology becoming more and more important from a sort of national security and just government level. So that was a really good interview from Senator Mark Warner on CNBC. Maybe now 2 months ago, and he was sort of asked about -- he's -- he chairs the Intel committee in the U.S., the Intelligence Community on the Senate. And they're sort of responsible for looking at strategic competition geopolitically, and they asked about what technologies you might have seen like with, for example, the Chips Bill was a consequence of the U.S. feeling like in the area of electronics and AI, we need to make sure we were being strategically competitive, in particular, with China. What other areas he was asked, we need to be strategically competitive and he said, synthetic biology and advanced energy. So in other words, batteries and solar and then synthetic biology, okay, right? And I'm going to touch a minute why I think that is, but you also see it is reflected in President Biden putting out an executive order also near the end of last year on strengthening the bio economy in the U.S. And I went to this event at the White House. Jake Sullivan was running it. You had the Head of the National Security Council. You had DepSecDef. You had the heads of USDA, heads of DOE, heads of NSF, and they're all there basically being told, you need to come back with what your plans are to make sure we're winning in the bio economy, okay? And so I think you should expect -- I've been in the field of synthetic biology for 20 years and just sort of got named back at MIT in the early 2000s. This is by far the most momentum I have seen in terms of the U.S. government taking the technology seriously. And I think part of that is strategic economic competition and part of that is COVID. It is a consequence -- it is an awareness of biosecurity and the impact of infectious disease on our national security being obvious. So those are sort of the 2 halves of this. We're happy to be sort of basically participating in both. So on the biosecurity side, I just want to highlight this. We had a great year this as well in 2022. So we're running a program with the CDC called Traveler-based Genomics Surveillance, where we basically -- you might have read about this in the news recently with all the flights in from China that are being with basically voluntary monitoring from nasal swabs as well as wastewater collected from the planes in order to look for new variants of COVID, right? So there's a big experiment being run in China right now given the scale of the outbreak going on, that's very likely to generate new variants, but we're not getting much data out of the country. That's why it's important to have these more global surveillance and efforts like what we have in the airports here so we can at least start to see data as early as we can in airports. And with this program, for example, we caught the first sequence cases of BA.2 and BA.3 back during the Omicron wave last year. We are expanding this globally. We're quite proud of that, working -- and this is -- we have MOUs or rolling airport programs and other things in these places, Africa Union, Qatar, Victoria, Rwanda, Botswana, Saudi Arabia, all publicly announced MOUs or programs. I'll just also mention it. We had a program with IARPA. This is the intelligence agencies version of DARPA called ENDAR that we've been doing the last 3 or 4 years, where the goal was look at a sequence of DNA and identify was it engineered or not, okay? And so this is something that is really not -- this is where I think it's an interesting interface between sort of public health and national security. Why would you want to know if something is engineered? Well, that's because you're looking to see if there was a deliberate misuse of biology versus something accidental or emerging from nature. And so I think those types of programs we can run the ENDAR, that's like radar for engineered biology. We can run that on these genomic sequences at these sort of radar sites globally. That's part of the idea. And the theory here is, this should ultimately look like satellites, right? We all dealt with a pretty inclement weather this week. I bet you brought a raincoat and an umbrella because you had warning before you showed up here in San Francisco that we should expect some bad weather. We don't have that for infectious disease right now, right? We find out 3 weeks after it's blowing up New York City, right? That's completely unacceptable. So I think we will see the build-out of global biosecurity, and I'm confident the United States will be the leader in bringing that out. Last point I want to make. I mostly talked about biopharma here today because this is JPMorgan, but if you look at our active programs in Q3, we had 85 of them, 28 were in biopharma. So we're actually doing a lot of work at Ginkgo in food and agriculture, industrial, consumer goods with large companies in those sectors, chemical industry, ag industry and so on. Biotech is not just about drugs, right? Biology can impact really any physical goods market, but we hear a lot about drugs. And then honestly, if you look at a bigger level, if you look at sort of capital markets in general, we hear a lot about information technology, right? Like just a lot of value you might have seen, for example, ChatGPT from Open AI. If you've had a chance to use this, you can sort of type in and give a query to ChatGPT and it will answer you. Sam Altman, the CEO of Open AI is a friend of mine. I think it's amazing what they've done, right? But so I went and I asked it a question and I was like ChatGPT, what is more important to society? atoms or bits, right? Is it all the physical stuff that we do in the economy or is it the information technology, right? What is it? And okay, I got this long-winded, atoms and bits are both important, but they play different roles and different implicates, dodging the question. ChatGPT is dodging the question. So then I was like, okay, ChatGPT, but if you had to pick one, what is more important to society, Atoms or bits. If I had to pick one, I would say that atoms are more important to society. This is because atoms are the building blocks of the physical world, and they are essential for our survival and well-being. They make the food we eat, the water we drink, the air we breathe, the materials we build our infrastructure with, homes, roads, bridges and power plants. Without atom, society would not be able to function. That's right, ChatGPT. And so if you look at what we're trying to do in synthetic biology, biology is the only programmable substrate. It runs on DNA code, but what it does, it doesn't move information around, it moves atoms around. It builds things. It makes our medicines, it makes our food. It cleans our water. It will be more important in the future. We're really proud that we're playing a role in building that technology here at Ginkgo. I'm happy all of you are participating in this as well in the biotech sector. We should grow the world we want to see. My e-mail is up there, if you're interested in any part of this, we'd love to talk to you. Thanks so much for your time.

Great. And as a reminder, if you do have a question, feel free to raise your hand and we can get a mic to you. Awesome. So just to start off with the pre announcement this morning. So you noted that you expect the total revenue to be in the expected range that you guided to earlier this year...

I insist you start by asking Anna Marie so that I can take a breath.

Perfect. So the press release also noted that foundry revenues are expected to be below that range. So can you just walk us through what happened there? It sounds like some of it was timing related. So how much of a slip from a timing perspective? And any details you can give us on that front?

Yes, absolutely. So as Jason alluded, business model, we take upfront revenues as we provide services to customers, and then we also earn downstream value as we deliver on those programs. The kind of core service revenue side of the business performed as expected. We did have some discrete milestones that we talked about in our last earnings call and the range of guidance we provided assumed that a certain set of those milestones were achieved. We did achieve some milestones in the quarter, but the mix was different. And so while we're still working towards those and remain confident that we'll achieve them, they did not land in the quarter as expected. And so that contributed to the announcement today.

And I would just say, this is sort of a general challenge with -- as I mentioned, when we sign up a new program, we get those 3 sorts of value. We get the near-term fees, we get long-term downstream value share, and we get the IP data growth value. And the downstream value share is great. It's high margin. This is, I believe, the milestones we talked about were -- maybe I'll make sure we don't say what we weren't supposed to say, but in any case, they come in as high margin. And so they're really great from the standpoint of the business, but they are often unpredictable exactly when they're going to happen because at the end of the day, they're typically out of our complete control. In other words, we don't run every part of the process that would trigger a milestone. Often the customer would need to either show a commercial demonstration for the ones we're currently working on. It tends to be a manufacturing demonstration that would really trigger it. And so it makes it a little harder for us to predict.

Perfect. And then obviously, it sounds like since you're able to hit that total revenue for the year, but then biosecurity performed better than expected. So can you just walk us through that biosecurity segment? And how are you thinking about this as COVID really becomes endemic?

Yes. So -- and I touched on this a little bit in the talk. We don't really see biosecurity as a COVID-specific thing in the long term. And I think you see this in the elevation of biology from a national security standpoint, and generally, it is -- countries are now aware that infectious disease is -- it can be nationally disruptive, right? I mean in my lifetime, I'd say, COVID was more economically and socially disruptive than 9/11 was, right? And so you have this awareness of that. And I think people in infectious disease, they -- have this like pandemic or it's called like panic and neglect cycle if you like talk to these like walks from that area. And they're like, well, yes, everyone got worried about SARS or Ebola. But then as soon as we beat it, you got over it. We didn't beat COVID, right? Like we got to annihilated by it, right? So that panic and neglect are not there this time, right? So you are seeing these systems getting built and people are aware, the next one is not COVID, right? It's obviously going to be something different. So -- but I think you still have this reality of the variants driving enough interest to build infrastructure. And then you have, I think, the national security concerns and public health concerns at a country level driving long-term persistent infrastructure. So I do think it will happen, but you never know. I mean it is getting built right now, though, in my opinion, I think with high likelihood.

So how do you think that Ginkgo will continue to play a role in that biosecurity industry as it continues to evolve? Will you be partnering with governments? You mentioned the bio manufacturing event at the White House. How should we think about that progression in Ginkgo being a part of it?

Yes. So for bio security, we see an opportunity, both in the U.S. and globally. So that's why you see us doing all this work trying to move this out internationally. That's actually not just -- obviously, it's great to have more business, but you need it international, right? Like you want -- like if you aren't monitoring outside of your borders, you're going to get surprised by biology, right? Like that's just the reality. Like biology does not respect borders. We saw that with COVID. So we do think there's an international business for us and a longtime U.S. business. I think you'll see us play in the surveillance area like I talked about, like monitoring for infectious disease. We're not a diagnostic company. I don't think you'll see us make big moves in diagnostics. We're not a therapeutic developer, but we'd love to enable therapeutics developers that want to develop things like more rapid vaccine or therapeutic development using our platform, that would be great. But we leave it to them to actually develop those assets from our business.

Helpful. Shifting over to cell program. So what are you seeing in terms of demand for Ginkgo cell programs as we head into '23?