Illumina, Inc. (NASDAQ:ILMN) Evercore ISI HEALTHCONx Conference November 28, 2018 11:00 PM ET
Jacquie Ross - VP, IR
Sam Samad - SVP and CFO
Ross Muken - Evercore ISI
Hello and welcome everyone. Thanks for joining us here at sort of our inaugural HEALTHCONx Conference here in Boston. As many of you know, I’m Ross Muken, I run the Life Science Tools & Diagnostic Product here at Evercore ISI. It’s my pleasure having with us a company I’ve known for a long time and an individual who I’ve been getting to know quite well over the last year or so, since joining Illumina, Sam Samad, who is our CFO; and then in the back, we have Jacquie, who all of you know does a fantastic job on IR. So Sam has some exciting news to start with. So I’d let you get through the formalities, so that we can get into some of the questions.
Yes, so the exciting news is the Safe Harbor. So I’ve been asked to remind you that my comments today could include forward-looking statements. You should refer to our SEC filings for a discussion of the risks and uncertainties that could cause our results to differ materially from our current expectations.
It is our intent that all forward-looking statements regarding our financial results and commercial activity made here in today’s presentation will be protected under the Private Securities Litigation Reform Act of 1995. Also the information discussed today is qualified in its entirety by the proxy statement that Pacific Biosciences will be filing with the SEC in the future. And Ross thanks for inviting me and thanks for having us here.
Very nicely done. So mainly since you gave us the disclosure on PacBio that would probably be a reasonable place to start. So I guess as you think about sort of the proposed transaction, obviously it gets you into long read, it seems like in their business they were starting to go through inflection, primarily [ph] and cost to where they think can be a clinical box. I guess, as you think about sort of where this fits into the Illumina family, what were the key kinds of attractions, and how you see sort of the complementarity there and given your presence in short read. Just give us the sort of flavor for how you are thinking about sort of that asset being brought hopefully into the Illumina family.
Sure. Yes, so just to reorient the group, and I’m sure all of you or most of you know by now that we announced earlier this month the acquisition of Pacific Biosciences for a price of approximately $1.2 billion or $8 a share and you know we’re really excited about this acquisition, really excited about this transition to or acquiring this long read technology.
As you said, there are definitely some complimentary applications, where long read is really important. Let me just clarify a few things and let me just give some background on this acquisition. First of all, we absolutely believe that short-read sequencing SBAs SBS-based sequencing will continue to lead the platform of choice based on costs, capability, accuracy. It’s going to be the -- continue to be the platform of choice for most applications.
Having said this, there are some really important complimentary benefits that long read technology brings. And it’s in those applications where you have you need to explain some higher repetitive parts of the genome, applications like pharmacogenomics, applications like functional genomics where a long read technology becomes really important as a complimentary technology to short read sequencing.
So it’s not an application that competes today with short read sequencing, but it’s very complimentary benefits to what Illumina has. And so if you think about the background of this, I mean there’s obviously something that has been -- we’ve been evaluating for sometimes. We’ve had our eyes on long read technology for sometime; this is not something that’s developed in the last two or three months and needless to say.
But the reason it became very important now is a couple of things. One is PacBio has made some pretty impressive progress in terms of the higher degree of accuracy that they have been able to establish with their technology that they have established now, consensus accuracy of 99% [ph] which for long read technology is as I said really impressive.
And at the same time, they have established the path to a $1000 genome as well. So their cost has also continued to come down. And when we think about the two technologies together, again for the complimentary benefits that this acquisition brings to us in certain applications that we don’t play today, we think we can first of all expand some of those markets, markets as I said pharmacogenomics, clinical microbiology, functional genomics where we bring Illumina’s expertise and you bring the long read technology of PacBio, we can expand some of those markets.
We can also accelerate clinical insights and discovery with the combination of short read and long read technology. And, we can also advance the existing development roadmaps of both companies because you’re bringing both innovation engines together, and you’re bringing two sets of – you’re bringing a lot of expertise in both companies really I would say the best out there in terms of what they bring in terms of scientific insight.
And, long read obviously sort of distinct from the short read technologies just sort of a different area. A lot of focus had been on planned animal and a number of different pieces. How do you think of sort of the competitive landscape today in long read, you had nanopore based approaches etcetera. How do you think sort of that will evolve and how do you think about some of those newer competitive elements in that market that have a slightly different technological approach?
Yes, so maybe step back first and talk about the overall market in long read, and then we can talk about some of those -- how what the competitive forces are so to speak. I mean, the overall market right now along read is approximately 700 million or so today, and PacBio expects this to grow to about two and a half billion in five years’ time. So a compounded growth rate of 30%. As I said, those are applications that we don’t plan today, but where we believe long read technology brings a very important complementary benefit.
If I think about competition, Ross, I mean, obviously there’s a – the biggest competitor out there in the long read technology space is Oxford Nanopore. And we know Oxford Nanopore; we’ve evaluated them in terms of the competitive or at least in terms of the progress that we’ve made. We’ve had investment in the past in Oxford Nanopore. And so, they bring some advantages to the table for sure. I think the one area that they are still improving and haven’t gotten to the PacBio I would say, a level of advancement here is accuracy. They have not established the same accuracy that PacBio has. But they are definitely an important player in the long read space.
So maybe pivoting the sort of the base business, right. So, NovaSeq is probably exceeded certainly from an investor perspective of our expectations. The box has been tremendously transformative to a number of markets. I guess, as you sort of look at the cadence and the type of customer adoption we’ve seen this year versus sort of maybe last year, how would you kind of compare and contrast the evolution of that box and then take us maybe more qualitatively not qualitatively into how you think this sort of next iteration of adoption is going to work, as some of these new markets that you’ve enabled start to really gain some momentum.
Sure. No that’s a really important question, and a really important point about NovaSeq is that we started this journey with NovaSeq with the intent of democratizing access to sequencing, and really expanding the customer pool that utilizes high throughput instruments. When it comes to NovaSeq, I think you will see that there is much more varied customer base and that has not happened by chance, that has happened by design, it’s happened by design in terms of how we’ve also launched the flow cells.
We started out; we launched the S2 flow cell. Then after that, in Q4 of 2017, we launched the S4 flow cell, which was more geared towards these really high throughput instruments, high throughput utilization customers. So, in some cases running X instruments, HiSeq X instruments.
And then after that we launched the S1, which is more geared towards the -- I would say HiSeq customers that are looking to convert from HiSeq running lower throughput work. When I say lower throughput still on the high throughput end of the spectrum, but lower than the S4 customers.
When we think about the level of evolution of those customers, I mean it’s still very early days. As we said at the end of 2017, only 15% of the 850 or so HiSeq X customers have actually taken on their first NovaSeq vessels back in at the end of 2017. So this is a little bit dated but you can tell that that was still very very early days in terms of customers actually even ordering a NovaSeq. I’m not even talking about converting fully to NovaSeq.
But when you think about today, you know there’s still, we’re still early in that journey. We’re still you know very few customers have upgraded completely to NovaSeq. And, that customer base is still very varied in terms of also what they are using with different flow cells.
In terms of looking forward, I mean we have the S4 [ph] flow cell also coming out. We expect it to come out at some point in the future, and that’s also going to be even lower throughput flow cell than the S1, so it will appeal to some of those lower throughput customers as well on the high throughput spectrum.
And so we continue to see that democratization depending on what kind of work you want to do, that you target certain flow cells. And the byproduct of this has been that this is one of the reasons that it’s been challenging for us to give a pull-through number, utilization number on NovaSeq because you have such a varied customer base, it’s appealing to many different customer types. And at this point in time, it’s very hard to give one number that captures the entire diversity of the customer base.
How should we think about the X conversions, because that’s obviously sort of a unique customer, a very high volume customer, again in terms of what happens there versus the traditional NovaSeq or the new adopters that will influence the consumer will pull-through, just help us think through where that base is and sort of contemplating a transition?
So of the 45 or so customers that have -- that are running HiSeq X, I would say very few of them going back to my previous comment have entirely upgraded to NovaSeq, very few of them. A lot of them have bought a NovaSeq, but that’s still early days. And consistent with the comments that we made in the past, we expect this to be a multi-year upgrade cycle where this is going to take time. The reason this is going to take time is again, it’s driven by customer preferences, driven by customer timing, customer transition. It’s not necessarily driven by the fact that hey we want our customers to upgrade. Now upgrade in this quarter or that quarter, it’s really entirely driven by the customers, and it’s driven by the customers because some have validated work flows that they are running, some have not started their particular work on NovaSeq or new projects on NovaSeq and haven’t validated work on NovaSeq.
Some are running a combination of HiSeq X and NovaSeq but then have not transitioned fully to NovaSeq. So it’s depending on where the customers are sitting in some of those workflows and samples that they’re running.
So maybe thinking about some of the new applications that are being enabled vis-à-vis sort of where the cost curve has come down. I guess, how are you thinking about some of the new areas that are driving sequencing consumable growth. I’m thinking about rare disease. I’m thinking of population based sequencing etcetera, some of the new oncology applications. How would you characterize kind of the growth in some of those areas and where we are and the size of those relevant opportunities comparative to each other?
Sure, yes. We’re seeing some really exciting vectors out there, that are driving that growth, that have driven that growth over the last few quarters, as we’ve talked about in terms of our sequencing consumables momentum. And I think you’ve hit on a couple of them. NIPT is one exciting area that we’ve seen whereby we have -- we have seen a lot of I would say expansion of an NIPT usage.
I mean, and in terms of coverage as well, where we have 95% of high risk NIPT covered, we have 43% of average risk NIPT coverage. So we’re seeing a bit of a migration, and a bit more focus around enhancing the coverage of average risk. But we believe that that’s also create a lot of momentum in the mid throughput space and mid throughput sequencing consumables.
And definitely our VeriSeq NIPT Solution that we have in Europe has also been a key vector of growth in the NIPT space. When you think about oncology and other key area, and other key vector of growth, that’s been really exciting. And so, the reimbursement I would say climate has really looked a lot more positive recently. You’ve got presence of payers like Anthem deciding to cover panels and stage four cancer in recurring metastatic lung cancer. And so there’s some really important advancements in terms of reimbursement that is also driving potential sequencing consumable usage.
You’ve got in the RUGD space that you talked about we’re seeing good private coverage and we’re seeing potential for also RUGD to migrate as part of standard-of-care for the undiagnosed disease. The recent decision by CMS to issue a price for whole genome sequencing is $5000 with an important development in that space.
You mentioned population sequencing and population genomics, that’s another area that’s been a key vector as well of future growth. I wouldn’t say its current growth, but a future growth, initiatives like, and I would say that’s been the most immediate in terms of revenue impact GeL and the collaboration with GeL towards sequencing 100,000 genomes which we expect will complete this year has been one vector. But there’s also a very exciting future initiatives that we expect in the future will drive sequencing consumables growth and potentially placements of interest on that.
Just on that, are you surprised at sort of the breadth of geographies or countries that are kind of moving forward with population based approaches, and is a little bit surprising until recently the U.S. maybe actually fell a little bit you could almost say behind, given how far ahead of the curve we’ve typically always been around these sort of progressive technologies. And do you think with all of us and some of what they are trying to do, and Helix [ph] partner of yours and what was going on in May prior to the individual leaving, we’re starting to see some action in the U.S but still other parts of the world might actually be slightly ahead of us and hopefully we may catch up.
Yeah. I’m not sure if surprise is the word. I mean, in some cases, honestly very pleased with the progress that we’re seeing outside the U.S. And the U.S. even though has lagged behind, I mean there’s been the Cancer Moonshot initiative was one area where they look at the population sequencing.
The other part is you’ve mentioned the all of us so that’s a really important development you know looking at sequencing and genome type taking a million samples over the course of the next 10 years. So that’s an important progress. I don’t want to diminish it at all. But yeah, to your point, and other parts of the world I think the key initiative has been the Genomics England initiative, which was announced a few years ago, which has really advanced towards the completion of that, those 100,000 samples which is now migrating into standard-of-care and into clinical practice for managing cancer and diagnosing a rare undiagnosed disease.
And yet, we’ve been really I would say pleased with the progress that we’ve seen in countries like Australia as an initiative. We’ve even talked about an initiative in Bangladesh, which is where we’re starting to see population genomics in some of the developing countries, understanding the benefits of sequencing at a population genomics level, and precision medicine in some of those countries. That, that’s going to be a huge, because it could be a precursor for other initiatives.
But, I think GeL and the NHS has been a good reference point for some of those initiatives to start looking at the work that they’ve done and say, okay, there’s some merit to this and maybe start to expand population genomics.
And I guess what do you see is the sort of next iteration or next phase of utilizing all this data that’s being driven from some of these PopSeq projects and eventually broadening out kind of the clinical applications that have in the market and bring it to a level where it really becomes a regular part of the standard-of-care for all babies born or all individuals that have certain types of diseases essentially that you could be more preventative toward if you were more informed earlier on. I guess, how do you see that kind of evolution laying out, because it does feel like the payer university within the U.S. a switch has kind of flipped more recently where they’re starting to understand. Wait a minute. There’s -- there might be an upfront expense, but the merits or the payoff I’m going to get from being more informed in my population even if it’s a small percentage of my population more than pays for it.
Yes, exactly and I think that’s the key, the key point, is how does this translates to health economic benefits to the fact of the knowledge that sequencing in medical practice and clinical practice will actually yield certain benefits both economically obviously clinically as well, that you know that these type of populations genomic studies will demonstrate. I think that’s a good precursor for this.
Again, I go back to the GeL because that’s the most advanced one, is the easiest to talk about, has been the fact that they’ve completed the 100,000 they’re in the midst of the interpretation stage and getting in and essentially analyzing all of this data from the 100,000 genomes. And it’s not completed yet, but they’re towards, they’re well towards completing that goal.
But now again as I said, they’re going into the next phase, where it becomes part of clinical practice. So I’d say that’s a good precursor for how some other health systems are going to look at what GeL has done and maybe follow that model. But I think the key point to keep in mind is what are the -- what are the clinical benefits that have been demonstrated and clinical insights. And also what are the health economic benefits that are demonstrated.
I mean, when you think about RUGD for instance, which is part of the scope of the GeL initiative and that diagnostic odyssey obviously clinically and the pain that these families go through, not just the children, but the families of the child that’s battled with this disease is a significant burdensome diagnostic odyssey, obviously healthwise, but also economically as well.
And when you can demonstrate proof points to that effect, that becomes really important and that I see is the next stage in terms of taking all of these insights and building them into clinical practice.
And you think it will be answer of the rare disease side possibly, where we see some of the early benefits of this is because you have so many patient groups that struggle to identify what disease is it truly that someone has or what is the possible treatment path where we might spend five, seven years prior with legacy technology to even make small progress with where with sequencing you could have something figured out in a child in the first three or six months of existence.
You’re absolutely right Ross. Yes. I mean, I think this is very compelling and you can see it in other areas like – obviously, oncology is another big area. But in rare and undiagnosed disease we have some really powerful examples, patient stories, where you have children that have been diagnosed and diagnosed and misdiagnosed and simple solution like giving them a daily dose of vitamin would have prevented a lot of complications that they could have had. And so the only way that was discovered was through sequencing.
And there is no more important way of demonstrating that than to tell some of these patient stories. And we have a trial which is unique HiSeq trial that have demonstrated also some of the benefits of -- health economic benefits and the benefits of diagnosing RUGD patients. And we’re working with a lot of, I would say, pacesetters or trendsetters out there like Rady Children’s Hospital in terms of looking at how they can diagnose, rare and undiagnosed disease in record time.
And so maybe just finishing with some of these emerging growth areas; you touched upon oncology, liquid biopsy particularly not just a diagnostic but also is a screening test, huge promise, I guess how are you thinking about progress made and some of the efforts there and kind of where we are in the evolution of that opportunity?
Yes. I mean, there's been some good progress made so far with both with GRAIL and also with Freenome and Guardant, so there’s a lot of players in that space. I think I wouldn’t characterize it necessarily as early days, but I would say there's still a lot of potential for growth in that as some of these studies also yield some of the results with GRAIL and others. And frankly with -- in terms of the benefit that Illumina can bring, I mean, I can't think of a better platform out there than NovaSeq for allowing some of that work to be done, but really high throughput, deep sequencing that needs to be done in the liquid biopsy space and bring insights to patients and potentially improve human health. There's no better platform out there, because the amount of sequencing that NovaSeq can do where it’s really not necessary for liquid biopsy. Its really important, it’s critical, becomes very very critical.
I just wanted to see, do we have any questions in the audience? All right. Maybe we just spent lot of time in talking about the high throughput side of things. Help us translate those, you’re obviously CFO to the financial side. So I guess give us a flavor for kind of where consumable growth has been at the high-end roughly and sort of the dynamics as we jump off into next year. I’m not asking obviously for next year guidance, but more in terms of what you need to do to kind of sustain the levels of growth we been used to seeing at least in the high-end portfolio?
So, yes, obviously I won’t go into the financials. We haven’t given 2019 guidance. So I can’t speak to specific percentages but this year has been a strong year for sequencing consumables growth driven by the introduction of NovaSeq, the utilization on NovaSeq, which by the way in terms of sequencing consumable utilization on NovaSeq and pull-through even though we haven’t given that number and we’re constantly asked about what that number is. But in Q2 we had a record pull-through on NovaSeq. In Q3 we had another record pull-through on NovaSeq. So, that's definitely, I would say NovaSeq has catalyze additional applications, additional sequencing consumable utilization in a lot of areas and that's been borne out with the growth that we’ve seen on consumable.
When we think about going forward, I mean, first of all let me talk directionally, but then let me talk about what’s driving that, when we expect to drive that direction? You know, directionally as we’ve said over the course of 2018 we would expect sustained growth on sequencing consumables over the next few periods as we continue to see NovaSeq consumables ramp up, potentially offset by HiSeq and HiSeq X consumables coming down as that transition continues which as I’ve said before, is a transition that we expect will take a number of years, not a number of months.
So, that’s sort of the overall momentum. Now that doesn't necessarily mean that every single quarter is going to be X percent of growth because you will see some choppiness in that depending on that consumer transition, but that’s the overall direction, overall trend. In terms of what's driving that? It's really without going through all of the areas that I mentioned before, but it’s going to be driven by a lot of these areas that I talked about before which are some really exciting growth vectors for us, oncology testing, oncology research, NIPT, rare and undiagnosed disease utilization and testing population genomics initiatives, although probably less so in 2019 maybe more so down the road because we expect those to ramp with a longer time frame.
And maybe talking the lower to mid throughput portfolio, obviously you touched upon NIPT before which is a key driver for NextSeq, but I guess how should we think about -- what has been outside of NIPT driving NextSeq, MiSeq. And then what do we’re seeing as if any thing you can kind of derive from the early HiSeq, obviously its small, but it is a new part of the fleet gives some new players early access or new access to sequencing as a whole. How would you kind of characterize what’s been driving demand at the low end of the portfolio?
So, maybe start with NextSeq, I’ll talk about it very quickly even though your question is more focused on the lower throughput. But NextSeq has really been workforce for us in the clinical space. And we’ve talked about consumable utilization on NextSeq as well and pull-through, and its been at the – its not at the high-end of our pull-through guidance range for NextSeq, sometimes exceeding that pull-through guidance range for NextSeq. So, we’ve seen very very solid growth on NextSeq and utilization there.
On the lower throughput, so what we call the desktop side of the instrument portfolio, so where you have MiSeq, MiniSeq and most recently HiSeq. We’ve had some interesting and really exciting introductions that we’ve had there. So on MiSeq, for instance, we’ve just recently received clearance for MiSeqDx in China which will help accelerate again MiSeq utilization or MiSeq usage for clinical practice and clinical insights and clinical discovery. We’ve recently announced new software for on the mid throughput side for NextSeq 550 for enhanced analysis tools for NextSeq.
Switching back to HiSeq and I know your question was around what are we seeing in the early days. First, let me talk about HiSeq in general. Our most acceptable benchtop instrument, our more accessible instrument period with the cost – capital cost of less than 20,000, we believe it can expand our customer base to more than 50,000 customers because it can appeal to those customers that either are sequencing today and using Illumina instruments it can appeal to customers that are not sequencing because they been outsourcing some of their work or it can appeal to customers that have never sequenced before that are starting to sequence for the first time.
So very accessible and expand the customer base. What we’re starting to see, I mean, we’ve been very pleased with the introduction of HiSeq with the shipments that we’ve seen, the orders that we've had for HiSeq. What we’re seeing in terms of early applications mostly has been customers as we expected are sequencing today have Illumina instruments and they are using HiSeq, so basically quality control certain lungs ahead of doing them on NovaSeq.
When you do runs on NovaSeq or a HiSeq X or HiSeq for that matter, I mean, they cost significant amount and you have a potential for error as you do these runs. And so doing them on an HiSeq will cost you in the hundreds of dollars and it will spare you a lot of the cost that goes with a failed drug for instance on NovaSeq or HiSeq. So those are the applications we’re starting to see, they will ultimately evolve into other applications as we start to penetrate those other customer segments that I talked about.
And maybe on the array side, there’s sort of been renaissance on in that business to see it growing as it is, is remarkable, given where we were a few years ago where people thought that business would be in decline. Could you kind of tease out sort of the base with sort of research part of the business from the consumer genomics side. And help us understand this sort of demand dynamics in each and really how much consumer is sort of driving some of the reacceleration versus the traditional applications?
Sure, yes. I mean, so arrays overall, yes, there’s a components to them which is some of the more mature applications I would call them which is agrigenomics and other application in the array space, but – and the other portion would be in the consumer space. And the consumer space would be to -- I would say types of revenue that we get. One part is basically the consumables and instruments that we sell to some of our customers that do consumer testing, that basically sell kits and do consumer testing, customers like 23andMe and other customers out there to do that and they utilized their own lab to do some of that work.
Then we have other array revenues that we get, which is more services array revenue, which is related to mostly to ancestry that sell consumer kit. They get the sample. They send it to us. We process it. We send it back to them. They sent to the consumer. They send the results to the consumer. And so that's sort of the component of that business. Obviously the agrigenomics part of the array business has been more mature, more I would say, consistent in terms of revenues. The consumer portion of that business has been going really rapidly. Last year -- last quarter I should say, we talked about array growth and we also add significant amount…
Yes, exactly, significant instrument placements for one particular customer ahead of the holiday season. So that sort of the dynamic in the array space. What we’ve seen going back to the first point that you mentioned is definitely, definitely an inflection in 2017 in terms of the number of samples that were processed in the microarray space in the consumer space, seven million samples which outpaced the ten previous years combined. So, it has been very very impressive in terms of growth. It’s been very resilient. And what we’re starting to see, Ross, as well is we’re starting to see other applications, and we can talk about that if you want. And we’re starting see also the transition. I’m not sure -- maybe transition is the bad word, but the expansion to outside the U.S. as well, and so we’re starting to see other countries as well that we’re seeing consumer testing.
And so as we think about the sustainability of consumer, Luke and I have done a lot of digging in recently, we met with a number the key players there, all of which your customers. The growth rates we’re hearing in terms of expectations for 2019 and beyond remain very elevated. So I guess as you think about sort of the evolution of that market initially as you said, it was sort of ancestry and maybe a little bit of sort of unique things around your own personal health. It seems like it's developing in tons of new areas. [Indiscernible], you can find your own color tartan and make socks out of that. I was buying my wife for the holidays something around sleep. There’s tons of new applications, obviously Helix has help develop that, but 23andMe as well. And then there's also now this push with the 23andMe and Glaxo collaboration around using consumer genomics to drive clinical trial enrollment in drug discovery. So I guess, as you think about the duration of that acceleration I guess given what the customers are saying you have more confidence now that that business will sort of stay, I mean that’s elevated as we've seen, but certainly in a much higher rate to kind of support array growth for the foreseeable future.
Yes. I mean, as I said, this business has been, definitely going through, I would say, pretty impressive growth phase. And to your point, I mean genealogy so far has been the killer app, right? Genealogy for ancestry for 23andMe has definitely been the key app that's driven some of this growth. But as we think about going forward you do have companies like Helix which is an affiliate that has an app marketplace where you can go and get sequence and sequence not based on arrays, but its based on excellent plus sequencing and then you can purchase different apps depending on what you want to get information on. Those apps could be genealogy or they could be lifestyle app, but they could also be health app. Helix just launched recently a collaboration app with mail which assesses health and disease risk. So that’s the new app that they have also on the marketplace.
Then you have other companies like 23andMe for instance which have gotten FDA approval to release or to work with consumers on pharmacogenomics results. And how – what the drug interactions might look like based on pharmacogenomics. You had -- again, I go back to the point about expansion outside the U.S. as well which is really been interesting to see. Companies in South Korea like Genoplan for instance that have partnered with insurance companies to release nutrigenomics results. You have companies like myDNA in Australia that also look at releasing pharmacogenomics result. You have WeGene in China that have recently just opened another lap to service Hong Kong and Southeast Asia and they have collaboration with us in China around setting up a lab for genealogy for 56 ethnic groups in China. So you’re seeing a lot of that expansion into other aspects, health being a key one of them, but things outside of health as well.
And may just thinking a little bit about that the P&L overall. We obviously had a period where there was some compressed margins given some of the investments that were coming in through GRAIL and Helix and we’ve sort of gone through that. PacBio will be a separate discussion. But as we think about this to underlying business and the operating leverage we’re expecting to get -- you obviously outperformed in the revenue side which is kind of drop down in the more material rate than many of us were looking for you raise your guidance all year. But it seems like the incremental margins underlying have been pretty steady in the 30s. Is that how we should be able to think about the business? Or is there anything for the unique that you've talked about at least on the investment side we should be aware of as we can jump off into next year?
Yes. It’s a good question and it captures a lot of things on the P&L, so I’ll try to sort of bring it together. You’re right. In terms of the operating margins we traditionally been in the low 30s, mid-30s and some of this is structural as well depending on. We used to report stock-based comp outside of operating margin. Now we reported inside operating margin, so creates a bit of – it doesn’t impact EPS but it creates some impact on operating margin. But traditionally our target has been I would say around 30% or 35% if you exclude stock-based compensation.
Our approach has been, listen, there are going to be years, if we’re launching a new platform and we’re making significant investments it might impact leverage I would say in a bit more of a transition, as we as we transition a lot of this investment that year, as we – and this is transient, this is not necessary structural. There might be certain points where we decide to make transformative investments and I emphasize transformative investment where it could impact leverage as well. Because we decide that this is an opportunity that's really important either because it brings new technology in, complementary technology or it essentially catalyzes the market, the part of the market that was -- that needs Illumina to catalyze. You talked about some of the initiatives that we’ve had with GRAIL and even Helix.
So, there could be some of those deviations that we see because of transformative investments that we make. Overall I would say we are definitely committed to leverage. We’re committed to our operating margin target that I mentioned earlier, but it could have transient, it could be impact in short time frame depending on that.
Could you just give us a little bit of color, I mean, the R&D budget is getting so large now, right? Just given so many of the areas, you've entered in the complexity of some of them. How you think about sort of what really is sort of underlying. I think one of those tough things when they analyze a company like Illumina, we can look at sort of what we know, but obviously you’re spending quite a lot of money. So there’s a lot we don't know that not just sort of iterations of the existing portfolio of how you think about architecting the budget, higher risk versus lower risk sort of investments in terms of some of these transformative than may be going on not to a degree to which GRAIL or something was but underlying. Just give us a flavor for the approach in philosophical kind of part of it?
Yes. Probably one of the toughest parts of my job, because we have so many exciting things going on, is how do you be able to make sure that you’re balancing what you need to return back to your shareholders, but also at the same time investing in your R&D pipeline which is so exciting and there's so much going on. So we've invested traditionally about 18% of our topline in R&D and that's definitely exceeding a lot of the comparable players out there or lot of players in the sector or just peer companies I would say.
And to your point I think you’ve captured it well, Ross, which is a portion of that is investments in R&D around our existing pipeline and the existing architectures that we have and how we can evolve this architecture. The $100 genome path that we talked about is a great example of that because we’ve said NovaSeq is going to be the architecture that’s going to get us to the $100 genome some day over the next few years and lot of the R&D that go towards evolving that architecture to get us to that space.
Another portion though is around the new parts of the pipeline or new innovations in the pipeline and how do we develop other platforms and innovations that that get us to also the next level, but did not necessarily on the same architectures that you have. And then you have another piece of the -- and then you have another piece of the pie which is also around what we call our research and technology development which looks at some much more I would say strategic R&D investments that are maybe more long-term in nature. So it’s compartmentalized that way. And my job is really to force us to be extremely disciplined on how we prioritize R&D. How we look at those things that are really going to be the most meaningful for our customers, for our patients that are most aligned to our mission of improving human health and be very focused on how we execute.
And I guess lastly, we have few seconds left. You typically have used the competitor conference to sort of announce major pieces just maybe next year this venue is close enough in time to that, and you can maybe spread around, but let’s just put that aside. As we think about sort of how you’re prioritizing just leading off of R&D different functions, I guess how are you thinking about sort of new platform kind of investment versus sort of new application versus software and some of the other piece. I guess what sort of top of mind for you right now as your thinking about the different areas of where the biggest needle movers can kind of come from or it’s a bit of mix of still everything?
Sure. Yes. I mean again it's an knowledge and commitment then I would say a discipline of making sure that we have -- we are carving out a portion of R&D to develop the architecture that we have because we have a commitment to reduce the cost of sequencing, and we’re going to work towards that commitment. But then we also have you know we allocated investments towards, as I said before, new platforms, new innovations that we’ll continue to do whether it’s in the high throughput, mid-throughput or low throughput space.
Excellent. Well, Sam, Jacquie, thanks for coming. You always make it exciting for us. So, always great to hear an update on the story and we look forward to hearing great things from you in 2019. Thanks everybody.
Thank you, Ross. Appreciate it. Thanks everyone.
End of Q&A