Illumina, Inc. (ILMN) Q2 2013 Earnings Report, Transcript and Summary

Rebecca Chambers - Senior Director, IR Jay T. Flatley - Chief Executive Officer, President and Director Marc A. Stapley - Chief Financial Officer and Senior Vice President Christian O. Henry - Senior Vice President and General Manager of Genomic Solutions

Tycho W. Peterson - JP Morgan Chase & Co, Research Division Douglas Schenkel - Cowen and Company, LLC, Research Division Derik De Bruin - BofA Merrill Lynch, Research Division Ross Muken - ISI Group Inc., Research Division Amanda Murphy - William Blair & Company L.L.C., Research Division Daniel Arias - UBS Investment Bank, Research Division William R. Quirk - Piper Jaffray Companies, Research Division Amit Bhalla - Citigroup Inc, Research Division Daniel Brennan - Morgan Stanley, Research Division Jonathan P. Groberg - Macquarie Research Isaac Ro - Goldman Sachs Group Inc., Research Division Daniel L. Leonard - Leerink Swann LLC, Research Division Zarak Khurshid - Wedbush Securities Inc., Research Division Peter Lawson - Mizuho Securities USA Inc., Research Division

Good day, ladies and gentlemen, and welcome to the Second Quarter 2013 Illumina Inc. Earnings Conference Call. My name is Ayisha, and I will be your coordinator for today's call. [Operator Instructions] As a reminder, this call is being recorded for replay purposes. I would now like to turn the conference over to your host for today, Ms. Rebecca Chambers. You may proceed, ma'am.

Thank you. Good afternoon, everyone, and welcome to our earnings call for the second quarter of fiscal 2013. During the call today, we will review the financial results released after the close of the market and offer commentary on our commercial activity. After which, we will host a question-and-answer session. If you have not had a chance to review the earnings release, it can be found in the Investor Relations section of our website at illumina.com. Participating today will be Jay Flatley, President and Chief Executive Officer; Marc Stapley, Senior Vice President and Chief Financial Officer; and Christian Henry, Senior Vice President and General Manager of our Genomic Solutions business. Jay will provide a brief update on the state of our business and Marc will review our second quarter financial results. This call is being recorded and the audio portion will be archived in the Investors section of our website. It is our intent that all forward-looking statements regarding our expected financial results and commercial activity made during today's call will be protected under the Private Securities Litigation Reform Act of 1995. Forward-looking statements are subject to risks and uncertainties. Actual events or results may differ materially from those projected or discussed. All forward-looking statements are based upon current information available and Illumina assumes no obligation to update these statements. To better understand the risks and uncertainties that could cause actual results to differ, we refer you to the documents that Illumina files with the Securities and Exchange Commission, including Illumina's most recent Forms 10-Q and 10-K. Before I turn the call over to Jay, I would like to let you know that we will participate in the Morgan Stanley Global Healthcare Conference in New York the week of September 9. For those of you unable to attend, we encourage you to listen to the webcast presentation, which will be available through the Investor Relations section of our website. With that, I will now turn the call over to Jay.

Thanks, Rebecca, and good afternoon, everyone. I'm pleased to report that Q2 was another exceptional quarter for the company. Our business demonstrated strong underlying trends in every geography in nearly all product lines. As a result, Q2 revenue grew 23% year-over-year to $346 million, our 7th consecutive quarter of sequential revenue growth. Total microarray revenue increased 3% year-over-year, due primarily to the impact of BlueGnome. Demand for arrays remain stable as the market trends toward large sample numbers with moderate content at a lower ASP per sample. In fact, Q2 is the third quarter in a row with Infinium orders exceeding 1 million samples due to demand for our Infinium, OmniExpressExome, Core Exome and Omni2.5 products. We will begin shipping a new custom onco array in the second half of this year to early access consortium members interested in understanding the genetic basis of cancer. The onco array is the next generation of the groundbreaking iCOGS custom array that was used by the collaborative oncological gene environment study to identify genetic variants related to breast, ovarian and prostate cancer. The onco array contains approximately 600,000 markers for use in a meta analysis of over 400,000 samples. We plan to offer this array to non-consortium customers in 2014. The consumer market continues to show a significant elasticity below the $100 price point, generating strong demand for gene mapping services and our iSelect custom arrays. For the second quarter in a row we received a large consumer order, deliverable over several years and we continue to project close to $50 million of consumer-related revenue in 2013. Turning now to our sequencing business, total sequencing revenue grew 33% year-over-year, driven by strong demand for consumables and HiSeq instruments. During the second quarter, sequencing consumables grew 26% over Q2 of last year as a…

Thanks, Jay. As Jay mentioned, this was another exceptional quarter for Illumina, and I'm pleased to report our operational and financial results, which includes second quarter revenue of $346 million, an increase of 23% year-over-year. Organic revenue grew 19% year-over-year as a result of continued strength in HiSeq instruments and growth in sequencing consumables. These results were based on strong global demand for our products. In Q2, revenue in the Americas grew 22% year-over-year, and European revenue increased 19% over the same period. In APAC, revenue grew 36% year-over-year due primarily to continued strength in Japan. Importantly, our revenue base is becoming ever more diverse, which has lessened to an extent our exposure to fluctuations in funding. Historically, approximately 80% of shipments were attributed to academic and government customers. Last year, we saw a strong demand from non-traditional customers, which led to approximately 30% of shipments coming from clinical, agricultural, commercial, hospital and translational customers. We have made great progress in further diversifying our revenue stream in the first half of 2013 as approximately 40% of our shipments were to customers in these applied markets. Instrument revenue for the second quarter was $95 million, an increase of 31% compared to the second quarter of 2012 due to upgrade revenue of significant year-over-year growth in HiSeq instruments. MiSeq unit shipments were slightly higher than last year's elevated level, which benefited from backlog clearing associated with the first few quarters in MiSeq placements. MiSeq orders grew sequentially, as well as year-over-year, with notable strength in APAC. Consumable revenue in the quarter was $215 million, an increase of 17% compared to the second quarter of 2012, primarily due to our growing installed base, higher demand for Sample Prep and sequencing consumables and BlueGnome. Consumable revenue represented 62% of total revenue, flat compared to…

[Operator Instructions] Your first question comes from the line of Tycho Peterson with JPMorgan. Tycho W. Peterson - JP Morgan Chase & Co, Research Division: First, just on the ALL deal. Can you talk about the average number of collaborations with Jmark [ph] Illumina and others? Can you maybe just talk about how you foresee those playing out going forward? And then also the up to $96 million, what does that assume for kind of milestones and other things going forward?

Yes, we are intending clearly, Tycho, to honor the obligations we have under the existing ALL agreements and those give to those third parties particular rights to the technology in specific market segments. None of those agreements, we think, preclude Illumina from doing exactly what is our goal with respect to the technology, which is very directly focused into the sequencing marketplace overall. In terms of the purchase price, it's a combination of an upfront and milestones. We're not, at this point, disclosing what those milestones are, what the amounts are. Tycho W. Peterson - JP Morgan Chase & Co, Research Division: Okay. And then just a quick follow up on diagnostic pricing. Can you talk about whether -- as your customers are commercializing tests, whether you're able to extract the better pricing? I guess, we've heard a little bit from the field about putting on supplementary pricing for some of the customers commercializing diagnostic tests? I'm just wondering if you can...

We are, and largely that comes through the what we call our TCG reagents channel. And as we began to get into the diagnostics market a few years ago, what we heard very clearly from those customers was that they required more in our reagent kits than we typically provided. That has to do with more extensive lot tracking, longer shelf life, the ability for us to continue to manufacture what would otherwise be obsolete products in the RUO space. As we layered these obligations that are appropriately demanded from us by those diagnostic customers, it clearly increases our costs of producing and holding those reagents. As a result of that, anybody who's using the TCG reagents pays a premium price to access those reagent sets and they get essentially premium capability and services associated with those kits.

Your next question comes from line of Doug Schenkel with Cowen and Co.

First question is during the quarter, you announced the CE Mark for a CF platform, a partnership with HistoGenetics in HLA testing and also plans to launch an HAI kit with pathogenic, 3 areas where multiplex tools have had a prominent role. Can you just walk us through why these areas? Are there more to come and do you have plans to work through the FDA process in some of these areas, especially CF where you did pursue the CE Mark in Europe?

We're looking to partner very broadly, Doug, across lots of areas of diagnostics. HLA is a very large market opportunity and we're engaged with multiple customers who want to use NexGen sequencing into that market. HistoGenetics was the first of potentially several that we might work with in that marketplace. You will see us continue to attempt to do this in many other areas, including infectious disease and cancer where we have lots and lots of different customers using the technology platform. So I think the partnering strategy there is working very well. What we will put through the FDA will be highly selective, and those are particular applications where we think having an IVD kit makes sense to be able to sell very broadly into labs around the globe. There may be cases where there's intellectual property in the partner we're working with that prevents us from doing that unless we specifically partner with that entity to put the product through the FDA jointly. In the case specifically of CF, we were able to do that on our own and we're now selling it in Europe under CE Mark broadly. And once we get it approved through the FDA in the U.S., we will sell it to all the labs broadly as an FDA-approved products here.

Okay. That's great. And then really, I guess, an expense question. Q2 was your first full quarter with Verinata as part of the company. R&D spending increased, I believe, it was $6.4 million sequentially. Is most of that Verinata and is this a good R&D run rate as we think about how to model the second half? And, I guess, more broadly, 2013 was positioned as a year where you were going to aggressively expand your commercial footprint. How is that effort progressing and how should we think about spend accordingly as we move into the second half?

Yes, I guess, I would say a material part of the increase in expenses was due to the incremental remaining part of the quarter of Verinata expenses. But you should expect through Q3 and Q4 a continued increase in R&D and SG&A spending. Obviously those spending increases will be, should I say, modulated appropriately. And so they're not going to be huge increases, but we're adding headcount in both areas, in R&D particularly, because we have a rich portfolio of products that we continue to work across and lots of demand for resources there. In the commercial side of the business, we're making great progress in adding sales people into the team, as well as new additions in marketing. Most of the impact of that on the revenue side will be seen toward the end of the year because as we add new sales people, it typically takes 3 to 6 months for them to fully come up to speed. But we're progressing well on the headcount additions in SG&A.

And, Doug, it's Marc. Just to be clear, I mean, the increase in expenses that Jay is referencing is implied in the updated guidance that we provided.

Your next question comes from the line of Derik De Bruin with Bank of America.

On the new 15G run for MiSeq, could you talk about a little bit of the technology that you got there? What's the cycle time improved to and where you are in the read lengths?

The improvements are largely due to new versions of chemistries, so we call it V3 chemistry. It increases the number of reads as well as the overall read length. The read length is [indiscernible] by 300. So we have great performance across read lengths. We also are increasing the number of tags to allow us to do more exome sequencing on that platform as we get the V3 kits into the marketplace.

So it's still running the same 5 minute cycle times? So a 15G run, then, is how long to take?

Cycle times, I believe, are going to speed up some.

I don't have that number right off the top of my head, Derik, but we have shrunk the cycle time down. Every time we try to extend read length, we try to shrink cycle time at the same time. And we were successful. I just don't know the exact number. I can follow-up with you after.

But the full run on a MiSeq with pair 300s will be approximately 60 hours.

Okay. Great. And then I just wanted to switch on another -- on technology -- another one, just could you talk about just the whole-genome sequencing market and where ASPs are on that and some commentary on volume? And also, just to follow through on that, where are you with your BGI contract?

Whole-genome sequencing, I guess, I'd say we continue to see migration of customers toward whole-genome sequencing away from exome and targeting to some extent, although exomes are continuing. HiSeq is clearly a fantastic platform for running exomes because you can multiplex them so highly and do it so inexpensively. As I said in the script, we had a record order received number for whole genomes in services, so we're continuing to see good progress there. Although that service ordering tends to be irregular in terms of both the pattern of orders, because they're large, and also in terms of delivery because we record the revenue when we complete the project and in some cases, when the customer actually accepts the data under some contracts. So great continued trends in whole-genome sequencing. As we continue to make improvements in technology, we're going to continue to make progress in the ability to do that ever more cheaply for our customers. With respect to BGI, we've completed an RUO agreement with BGI, and so that's been in place for some time, a month or 2. We are still working on a clinical agreement with them that's incomplete at the moment.

Your next question comes from the line of Ross Muken with ISI.

Can you give us maybe on Verinata, a little bit of color on kind of what the volume ramp has looked like? Obviously, we've seen what the key player in that market's done but just to kind of get a sense for obviously, it's been exponential growth as an industry overall. Trying to get a sense for how you feel like you're sharing -- faring from a share perspective and how you're expecting that to accelerate in the second half with kind of the efforts of your partner?

Yes, I mean, I guess, we're going to be cautious about disclosing volumes on Verinata, particularly because, as you know, we play broadly in this market and we supply to all of the major companies that are providing technology into this space. We are doing well, as I mentioned. We think both our covered lives and contracted lives numbers are progressing quite nicely, so we're doing very, very well there. We've continued to improve the technology that's being used by Verinata. We've put more sequencers in there automating the Sample Prep around the technology to be able to handle higher sample volumes. I think the reimbursement rates have exceeded our expectation, probably, in terms of where those numbers are coming in. So really, pretty positive trends across the board there.

Great. And maybe talk a bit about what Christian's been doing on the BD side, I mean ALL, really interesting little asset. It seems like you guys are being a lot more active, trying to fill-in all the various areas around the existing portfolio with sort of exciting new technologies. Maybe talk a bit about what the pipeline looks like? What areas, maybe not giving away specific targets, but just generally big picture areas that kind of get you the most excited and where you see kind of the best kind of emerging technologies maybe developing on the forefront?

Yes, we have been quite active in M&A and I think you'll continue to see us do some of these tuck-in type of acquisitions, fewer of the moderate to large ones and some continued rate of these relatively small ones. We are looking at technology pieces that accelerate the development of our future platforms, obviously our platform plans go out 3 to 5 years, and we have really great expectations about where the technology is going to head and how we'll be able to improve overall system integration, reducing hands-on time, making this more foolproof so that sequencing can become ever more prevalent in the clinic and used by less and less sophisticated users. So you can imagine, if you think about the entire workflow of sequencing from the earliest parts of the Sample Prep, all the way through informatics, we're looking to fill in any bits and pieces along that pipeline that we don't currently have. Some of that clearly is being done internally through our own research and development efforts but if we have the opportunity to go get it quickly outside at a reasonable price, we will do that.

And I also think, Jay, we should recognize our corp dev group, Nick Naclerio. The ALL transaction was complicated in the sense that there were a lot of different relationships to work through and they stuck with it and it's going to be a real fantastic asset for us. I'm really excited about it.

Your next question comes from the line of Amanda Murphy with William Blair. Amanda Murphy - William Blair & Company L.L.C., Research Division: So I had a question on the ordered arrays. I'm curious, from our discussions, it seems like your customers are pretty excited about that in terms of what it might do for throughput. So curious if you can provide a perspective on timing and how to think about what throughput improvements there might be relative to sort of overall demand for capacity at this point in the sequencing phase?

Yes. I think we're on-track, Amanda, to hit the targets that we set out earlier in the year, and that's to have the first ordered arrays in the hands of our customers before year-end in an early access mode. We think they'll probably become commercial in the very earliest part of 2014. We will deploy ordered arrays selectively in particular kits, where those applications demand high throughput. So not every one of our kits is going to get ordered arrays. Maybe ultimately, they will all move in that direction, but in the near term, we're picking off the most important application areas and applications in general in terms of read length and reads, both that need the ordered array technology. The performance improvements you'll get from ordered arrays will depend on the architecture you deploy it on. In almost all cases, we will get the advantage of more reads, but in our lower resolution optical systems, the density you'll be able to put down, lay down in terms of ordered arrays will be more limited than in the higher resolution systems, which obviously, is the HiSeq at the very high end. So HiSeq will get the greatest advantage, MiSeq will get a lesser advantage. Amanda Murphy - William Blair & Company L.L.C., Research Division: Got it. And then just I guess, I don't know if you've talked about pricing for ordered arrays yet, or maybe you could provide something there, if you're giving that data? And then also just generally, how you're thinking about pricing for HiSeq consumables, I guess, generally sequencing consumables going forward?

Yes. We have not disclosed any pricing on ordered arrays yet. In fact, we haven't even had our pricing committee meeting. So we don't actually know what the prices are and we're waiting to see what the final performance looks like off these arrays before we go into sort of the pricing phase, which will happen here in the back half of 2013. I'm sorry, the second part of your question, was what... Amanda Murphy - William Blair & Company L.L.C., Research Division: I guess, [indiscernible] thinking about you've had a few price increases on the consumable for sequencing, is that something you plan on continuing?

Yes, I think in general, our intention is to have moderate price increases in consumable kits on an annualized basis. And then we will deliver significant changes in value to customers when we have technology shifts, either in major changes in the kits or modifications in the platforms. And so I think if you think about what the pricing curve looks like, it will probably be one that will have some plateaus in it and then there will be some significant drops and then another plateau and then additional drops. And so it may not be sort of monotonically decreasing, it may be more digital, the way you might think about it.

Your next question comes from the line of Dan Arias with UBS.

Jay, you mentioned the work that the Mercy General folks are doing in pediatrics, which when they talked about it, sounded like a pretty excellent application. But it also seems fairly cutting-edge. So, I guess, how quickly do you think we'd see an acceleration in that type of quick turnaround work in the hospital setting?

I think the 2500 is a great example of how important these rapid turnaround capabilities are and what we talked about in the script, I think is really exemplary of what customers think about the ability to both manage their Qs on the one hand but also in clinical applications, get data back to the treating physician. We're working with lots and lots of groups now where they really want to get answers quite quickly, and it's very important in areas like infectious disease, transplants in HLA, less important perhaps in cancers, at least in the bulk of the cancer market where treatment decisions usually take a couple of weeks, and you have a little bit more leeway. But in almost every case, the patient wants to know as quickly as possible. And so whether you're doing an NIPT test or even in the cancer example, turnaround time matters. And so I think the technology we use with Stephen Kingsmore is largely technology we have today. So it's really just a packaging up. There's no brand new invention required to get to that sub 24-hour turnaround. It really has to do with sort of packaging up sample to answer sort of complete workflow in deploying that fully commercially to get it. There's really no breakthroughs required now.

Got it. And then maybe just a follow-up on the whole-genome question, can you talk a little bit about the IGN genomes that you are shipping, what kind of customers are those going to and I guess, what kind of an appetite is there from the clinically focused crowd?

Well, it continues to be a relatively diverse customer base. Some from pharma companies, a couple of these big projects like the VA project. I think you're going to see more and more from population-based sequencing programs. And if I were to turn the clock forward a year or 18 months, I would think that you're going to have more and more from these large programs. Some of those will wind up with local installations of HiSeqs, but some of them may, particularly in the earlier phases, use outsource services to get that work done. So I think the trend there is a good one. I still think that, as I said probably 2 years ago, that the outsource part of whole human genome sequencing is going to be an important but not significant part of the overall dollars in the marketplace, sort of in the 5% to 10% range, ultimately.

Your next question comes from the line of Bill Quirk with Piper Jaffray.

A couple of questions for me. First off, thanks for the disclosure around the ALL transaction. Just, I guess, the last housekeeping question here, what is that going to add in terms of revenue to 2013?

The revenue is insignificant, Bill, so it's not measurable.

Okay. Got it.

In terms of our overall revenue, at least.

Willing to take a shot at what we should expect in 2014? We certainly recognize, Jay, that due to some of the collaborations it obviously will ramp up once some of those products get commercial. But, I guess, how should we think about this pacing over the next couple of years?

Yes. I wouldn't think about the ALL revenue being standalone revenue. It's going to be so tightly integrated into what we're delivering as part of our product mix, that it will be indistinguishable into 2014. Yes, there'll be some revenue from the partnership projects, but those are, frankly, at this point, still speculative and we don't control the distribution channel there, so we can't really count on that revenue.

Got it. And then as a follow-up, and maybe a little bit bigger picture question here, but we haven't seen any material expansions of HiSeq in terms of throughput since the D3 launch a couple of years ago. So how should we be thinking about the future evolution here? Perhaps just HiSeq is one category versus some of your pipeline technologies?

Well, we have, as we said many times before, lots of degrees of freedom in terms of where we move the platforms in the SBS chemistry. And the important thing we did over the past 18 months with HiSeq was to accelerate the time of the runs, and that's really what customers wanted, was the most important new feature and new capability. It was frankly less important that we moved the total output up. And in fact, HiSeq now if you're doing anything other than whole human genomes, you need to multiplex lots and lots of samples. And so other than that segment of the market, it doesn't make sense to continue to push it too far too fast. As the market continues to go more and more toward whole human genome sequencing, clearly we'll have the capability to put higher output technologies into the market and continue to drive down that cost per whole human genome. So we're working across all those fronts, but we've prioritized speed of the runs, and that's why you saw the 2500 and that's what we're really going to focus on here in the next kits that's coming out on HiSeq before the end of the year.

Your next question comes from line of Amit Bhalla with Citi Group.

Jay, I just wanted to know if you could just talk a little bit about reagent consumption between the 2500 and the 2000 now that the 2500 seems to have been gaining some good traction? And also, can you give us a sense of utilization rates amongst your clinical customers versus your traditional academic customers from the reagent side?

Yes. I mean, I don't have the breakdown by system type in front of me of reagents. They're not radically different. So it's not something where we're seeing a 2x or even a 50% difference whether a customer has a 2500 or not. And even if they have a 2500, they probably run it half the time on average, for standard mode. And so I don't think that distinction is one that we spend a lot of time trying to focus on. I'm sorry, the second part of your question was on...

Yes, it was on the clinical customers. Because now that your customer base is also diversifying, and I had one quick follow up after that.

I think in the clinical markets, what we see is a sort of a bimodal distribution of reagent usage. You'll have customers initially buy the systems, whether it's a HiSeq or MiSeq, and use it in a development mode to begin to put their assay together. And during that mode, the consumption often isn't very high. And they're trying to sort of sort out exactly what genes to include in the panel, for example. Very often, they working on the informatics to improve the data analysis so they don't even need to run the system during that phase. And then you have the other customers who've gotten over that initial hurdle and are beginning to actually deploy the technology. And when you look at that customer base, the utilization rates are extraordinarily high, and that applies both to MiSeqs and to HiSeqs. So you find them in both camps, and then the net of that effect, of course, is that it tends to average out in the middle somewhere.

And, I mean, I would say another factor that plays into that is the amount of Sample Prep and the complexity of Sample Preparation that goes into that as well that, that can also create some kind of bimodal distribution in that population.

And then just a quick follow-up on Moleculo as a kit-able product. I think you said in the prepared comments that you plan to roll that out in the coming quarters. It sounds like there's a couple of gaming factors there. I thought it was more of a kind of end of 2013 rollout. So can you maybe just talk about Moleculo as a kit-able solution?

Yes, I mean, that remains our goal. We expect to have at least an early access before the end of the year. It may slip a little bit into 2014, depending upon how the product development goes here over the next quarter or 2. The challenges here are around areas like informatics. So we have to really work on optimizing both the ease of use for the front end Sample Prep part of this. But also, the compute time on the back end is important. And what we don't want to do is put a commercial product out there where the customers get overwhelmed by the computing requirements necessary to phase a genome or to get the synthetic long reads. And so there's a significant amount of work we're doing on that part of it as well, which is sort of unrelated to the actual chemistry.

And of course, Amit, by running it in our services lab, customers are getting experience with the data and what it can do in the different kinds of applications, but we're also getting experience on how to optimize the workflows. Jay said the workflow, making it a very simple workflow is a key ingredient to making it successful to the market. And as Jay said, early access around the end of the year is what we're shooting for right now.

And, that's why we were able to deploy it in the services environment first is because the simplicity of the workflow is less important. We can obviously train our internal team and they know how to run this quite well. It's different if you start offering it broadly to any customer who wants to buy it.

Your next question comes from the line of Dan Brennan with Morgan Stanley.

First question is just on the competitive environment. Jay, I know you mentioned it in the prepared remarks, but just wondering kind of given the slower pace of improvement in product introduction from your key competitor, just wondering if you could maybe give a little bit more color about what type of impact you've seen on your business from maybe a demand and/or pricing perspective?

Well, I guess, we break our market, really, into the 2 segments, the higher throughput segment and the desktop segment. And clearly, while maybe a year or 2 ago, there was some marketing of the competitive technology as potential replacement for HiSeq, I think that has tended to diminish quite significantly over that period of time. And so I think we're competitively in great shape with respect to HiSeq and particularly with the 2500 there. The desktop market remains competitive. I mean, they're a very capable and aggressive competitor and so we fight for all those orders. But I think we're continuing to win the vast majority of those where we're present. And again in the script, we said that, that number is at least 80%. And we're working on the coverage to make sure that we get into these other markets and geographies and application areas where we're not historically as strong.

Okay, great. And then maybe to follow-up on a separate topic, well, your nonacademic mix is growing rapidly, still be interested in just to get an update on what type of growth you saw this quarter from the U.S. academic and maybe the European academic customer base? And how we should think about that growth maybe as the NAH budget flattens out as the new fiscal year approaches?

Yes. I mean, we're continuing to see growth in the academic market. It's obviously largely driven by expansion of applications coming on the sequencing and growth from more dollars being allocated towards sequencing, and it's burdened by the fact that there's less total dollars to go around. And so while there's continued angst about the reduction in the NAH budget and we would love to see it increase as much as anybody else would, academically, we continue to do well. So I think that it remains a positive for us if all of a sudden the NAH budget went up 3% instead of down 5%, we'd do even better in those academic segments, but I think we're more than holding our own there.

Your next question comes from the line of Jon Groberg with Macquarie.

So just a couple of questions on sequencing. I think you said in the script that instruments were up 37%. And I'm just curious, I think you also said that 85% of the installed base had now upgraded to the 2500. So I'm just thinking about the second half of the year, how you're thinking about HiSeq demand relative to the first half of the year and MiSeq demand, I guess, relative to the first half of the year. It sounded like maybe MiSeq was picking up again a little bit?

Just to clarify, Jon, what we said was that of the people who ordered upgrades of the 2000 and 2500, 85% of those orders have been installed. So that's quite different.

And the instrument revenue was up 31%.

31%, sorry I was trying to write the notes down. I guess, I mean the question is still the same, just big picture, how -- I think there had been some effort to maybe try and do a little bit better job of pacing HiSeq so you didn't have the kind of the end of '11 event like it happened before. So I don't know, just kind of how you're thinking about HiSeq in the second half versus what you've seen here in the first half of the year?

I mean, I guess, as we look forward in our pipeline, which is pretty firm 6 months out and reasonably good visibility 9 months out, we continue to see a rich demand for HiSeqs coming from emergence of lots of new types of applications running on the platform. Clearly, we're seeing an increased fraction of the HiSeq market going into clinical kinds of applications and non-traditional applications. We think that's going to continue. In terms of the desktop market, that continues to do well in addition. So I don't think we're here predicting that we're going to fall off any kind of cliff that we see on the horizon. Things feel pretty good to us at this point. I think our guidance reflects the appropriate balance between upside opportunities and downside risk.

Let me just clarify. And when I say instruments are up 31%, that was total instruments. You'll note on sequencing instruments being up 37% was great.

And then, I guess, the sequencing follow up, if a little surprise sequentially that services will be down just because that includes Verinata, if I'm not mistaken, and that would've been the kind of first full quarter there of Verinata. So was there any meaningful downshift in the other services that we should be aware of, just the lumpiness that you were talking about? Just any insight there?

Yes. I mean, there's certainly some lumpiness in the services number. We had one significant services contract that slipped out of the quarter into Q3 that was pretty large in terms of revenue altogether. And it will be recorded in Q3. It was simply a customer acceptance question. Had to go through a third-party on the way to customer acceptance. And so I think that was the single biggest factor in the services line. But we did have really strong contract revenue in the third quarter. So that was very positive and then we had the full addition of Verinata on top of it.

Your next question comes from the line of Isaac Ro with Goldman Sachs.

Just a follow-up on sequester. I appreciate the comments you made earlier. I'm wondering if you could maybe touch on what's making your guidance for the balance of this year in the areas of the business where you have exposure? And the reason I ask is we're obviously still in the early months of settling in on the new run rate here. So I'm just trying to figure out how you guys are trying to handicap the spending behavior for your academic customers?

Yes. It's clearly an uncertain time there and there's a wide range of potential outcomes. I mean, we're clearly watching what's happening with respect to the budget debates that are starting to kick into gear here in the third quarter. Nobody has a clue really, what's going to happen with that and I think the chances of us avoiding a continuing resolution are nearly 0. So I think by the time we get to Q4, we're going to be under some new continuing resolution. So what we've done is we've modeled a broad range of sort of down sign funding scenarios, and then we've modeled what we think are a set of upside opportunities that the company has in terms of large deals and strength across the product line, funding, stimulus in Japan, sort of solid performance in Europe with the new budget structure going into place in Europe. And we've done our best to balance that out in our overall guidance. And so it, I think, is a number that we feel relatively confident we can deliver with that set of uncertainties.

Great. And then just a follow-up for Marc on the financials. With regard to the step up in SG&A the last couple of quarters, can you maybe delineate a little bit between what there is core to the business from the distribution initiative that Jay mentioned and then maybe what is incremental with regards to acquisitions? And then just lastly on tax. Tax rate, you guys have done a great job there. Any material opportunities left to continue lowering the tax rate?

Yes, Isaac. So on the SG&A spending, I mean, I wouldn't want to be convoluting into its individual pieces, but there are really a couple of things driving that and you listed them. I mean, we've got the whole quarter now of Verinata in SG&A, obviously, which includes things like legal expenses, the sales, the sales force that Verinata has and so on. And then regular G&A. We got some of our own process improvement projects taking place that we've talked about previously on these calls, and so we've had a step up in spending in those projects in the quarter. Those are really the key elements. I'm trying to think if there's anything else, worth calling out. You're going to see more of that as we go into, as Jay mentioned earlier, as we go into the second half, and you're [indiscernible] do more of that process spending, as well as adding more to our commercial infrastructure. So our commercial sales force and so on. On the tax rate, thanks for pointing that out. I think both the operational team that is involved in moving our manufacturing to the appropriate locations plus our finance team has done a really good job in managing our exposure from that standpoint. We're now halfway through the year so we've got a couple of quarters behind us, which is why we've been able to take our guidance to about 30% for the year, and we've done that through a combination of moving more of our manufacturing to our plant in Singapore and other operational factors as well, and we'll continue to do more of that. There might be some more headroom. We do think there will be. We have been targeting to trying to get that tax on an aggregate basis for the year down to the high 20s.

Just remember, Isaac, next year we won't have the benefit of the double R&D tax credit.

Right.

That's correct.

Your next question comes from the line of Dan Leonard with Leerink Swann.

First off, the number you gave, the 45% of your HiSeq shipments were into commercial and clinical applications. How does that compare to prior periods, if you've been tracking that?

It's higher. I mean, the trend line there is very strong because in overall the HiSeq business is doing better than we expected if we were to look back a few quarters, and I think the incremental improvement is coming from nonacademic markets.

Okay. And then for my follow-up, Jay, have you seen any impact on your order funnel from the Supreme Court gene patent ruling, presumably any impact of diagnostic labs wanting to get into businesses they were previously precluded from?

Not yet. There's a number of them who've announced diagnostic tests, but I don't think any of them are actually at the point where they've deployed those. They've announced that they're going to deploy them. And clearly, there's new litigation on top of those announcements that I think will be ruled on relatively quickly in terms of whether there's potential injunctive relief here. But no one to my knowledge has ordered new platforms, for example, to sort of host those new tests.

Your next question comes from the line of Zarak Khurshid with Wedbush Securities.

With respect to ALL, can you give us a flavor for how the technology might be commercialized? Is it something that would be integrated directly into a MiSeq or HiSeq or a future platform?

Well, I can generally answer that. Obviously, we don't talk specifically about our developments until we're ready to nearly launch them, if not launch them. But if you can imagine that the ALL technology could be used extensively in the Sample Prep process. And the ability to use smaller amounts of reagents impacts the cost. In particular, it influences ease-of-use because the customer gets rid of lots pipe hitting [ph] steps and multiple operations that are needed in the existing Sample Prep environment. For our high throughput customers, they're largely doing that on robotic platforms of various sorts. But for low throughput customers, it's more of a manual operation and we have the, potentially, the technology to really make that go way largely. If you fast-forward some number of years, then clearly, this gives us an interesting technology to begin to increasingly integrate with the sequencing itself. And over time, you've seen us pull more operation on the [indiscernible] and you'll see us continue to move in that direction of [indiscernible] Sample Prep to be more decoupled in some cases, in some [indiscernible] great core technology to begin to work on that problem.

Interesting. And then just kind of a high level question about the diagnostics business. How would you compare the relative sizes and growth rates of the various areas like cancer versus total prenatal versus consumer or other kind of growing segments of that business?

Well, if you look back a couple of years, consumer was clearly at the head of the pack in terms of the speed of growth and actual revenue. The last two years, it's clearly been eclipsed by what's happened in NIPT. That market has evolved faster than I think anyone could have ever imagined. And I'm hearing just stories all the time. We were out on a road trip talking to investors and we ran into multiple people who had actually had firsthand experience with the test itself, not, sorry, the Verinata version, but of the general NIPT testing. So the awareness level is spreading at rates that are unprecedented here. So that market is large and growing quite fast. I think if you look forward 3 years, 4 years, it will probably be overwhelmed by what's going to happen in the cancer market. And that's because the number of opportunities you have to sequence cancer patients is large. There will be screening applications in cancer as well as monitoring applications. And globally, it's going to be a vastly bigger market.

Your next question comes from the line of Peter Lawson with Mizuho.

Jay, just around the ALL acquisition, are you going to be expanding upon their diagnostic assays?

Well, we don't have any specific plans to announce around that now. I guess what I can say in general is that in most of the areas of the diagnostics market, our plan is to partner and to work with third parties who have intellectual property in particular areas or have know-how or distribution in particular markets. You will not see us broadly take as many diagnostic tests directly to the end users or to the physician, that's not generally our business model.

And, of course, ALL has already got some pretty solid collaborations that we will continue to support and fulfill all of our obligations there to enable them to be successful.

And on the onco array, just as you commented around the cancer market, is that going for FDA approval and does that cannibalize the sequencing panels you're generating?

We don't have any current plan to put that through the FDA. We will put it in the market broadly in '14 after selling it sort of on a propriety basis to this consortium group in '13. We'll clearly monitor its usage and how customers -- sort of how they apply it and what kinds of clinical environments it gets used. So we may change our minds about that at some point. But I'd say the probability of putting that through the FDA is relatively low.

There are no further questions in the queue at this time. I would now like to turn the call over to Rebecca Chambers for closing remarks.

Thank you. As a reminder, a replay of this call will be available as a webcast in the Investor section of our website, as well as through the dial-in instructions contained in today's earnings release. Thank you for joining us today. This concludes our call, and we look forward to our next update following the close of the third fiscal quarter.

Thank you for your participation in today's conference. This concludes the presentation. You may now disconnect. Have a great day.