Pacific Biosciences of California Inc (PACB) 2016 Q3 法說會逐字稿

Good day, ladies and gentlemen, and welcome to Pac Biosciences third quarter 2016 earnings conference call. (Operator Instructions) As a reminder, this conference call may be recorded.

I would now like to turn the conference over to Ms. Trevin Rard. Ms. Rard, you may begin.

Thank you. Good afternoon, and welcome to the Pacific Biosciences third-quarter 2016 conference call. Earlier today, we issued a press release outlining the financial results we'll be discussing on today's call, a copy of which is available on the Investors section of our website at www.pacb.com, or alternatively as furnished on the Form 10-Q, available on the Securities and Exchange Commission website at www.sec.gov.

With me today are Mike Hunkapiller, our Chief Executive Officer; Susan Barnes, our Chief Financial Officer; and Ben Gong, our Vice President of Finance and Treasurer.

Before we begin, I would like to remind you that on today's call we may be making forward-looking statements including plans and expectations relating to our financial projections, products, and other future events. You should not place undue reliance on forward-looking statements because they are subject to assumptions, risks, and uncertainties and may differ materially from the actual results.

These risks and uncertainties are more fully described in our Securities and Exchange Commission filings, including our most recently-filed report on Form 10-Q. Pacific Biosciences undertakes no obligation to update forward-looking statements.

In addition, please note that today's call is being recorded and will be available for audio replay on the Investors section of our website shortly after the call. Investors electing to use the audio replay are cautioned that forward-looking statements made on today's call may differ or change materially after the completion of the live call.

With that, I would now like to turn the call over Mike.

Thanks, Trevin. Good afternoon, and thank you for joining us today. We are pleased with our third-quarter results and our continued progress in driving growth in our business. Highlights of our Q3 financial results are as follows.

We generated $25.1 million in total revenue for the third quarter, up 80% compared with $13.9 million in Q3 of 2015. Excluding contractual revenue related to our Roche agreement, revenue grew by over 100%.

Instrument revenue increased more than 400%, from $2.2 million last year to $11.5 million this year as a result of our ramp-up of Sequel instrument shipments. We shipped and installed more than 30 Sequel instruments during the third quarter, bringing our install base of Sequel Systems to over 75.

Consumable revenue for the third quarter was $6.5 million, up 22% from $5.4 million recorded in Q3 2015. Consumable revenue for the quarter primarily reflected usage of RS II systems. However, we are starting to ship more consumables to Sequel customers as our supply of Sequel SMRT sales has increased.

Our gross margin for the quarter was 50%, consistent with the gross margin we have generated throughout this year, and up from 47% from Q3 2015.

We received orders for 20 PacBio instruments during the third quarter. At the end of Q3, our backlog was over 40 systems. New orders were somewhat fewer than we had expected to book in the quarter. As reported in our second-quarter call, we believe that limited Sequel SMRT sales supply hindered our sales of Sequel instruments through the first half of the year as prospective customers postponed their purchases while waiting for substantial data generation from the Sequel Systems installed with our early customers.

Although we announced unrestricted SMRT sale shipments in the third quarter, usage ramp was slower than we had hoped, partly due to the impact of summer vacation time, delaying the production of reference data.

Moreover, although those early customers focused on targeted sequencing applications were very successful, the large amount of DNA required for loading high molecular weight DNA sequencing libraries on the Sequel SMRT cells further slowed those interested in genome-wide de novo assembly or structural variant, or SV, analysis.

Last week, we began broad commercial release of new sample loading reagents and protocols that dramatically reduce, by a factor of more than 50, the amount of these high molecular weight DNA libraries required for optimal loading onto the Sequel SMRT cells.

Moreover, these changes should allow successful loading of even longer DNA libraries than those that customers can use with their earlier RS II sequencer, a capability that should enable even longer read lengths, important for assembly and SV analysis.

In early and mid-October, we released these upgrades to several beta sites that confirmed the loading performance gains. These customers could generate five to seven gigabases of output per SMRT cell, with read length in 50s of 15 to 20 kilobases on bacterial, plant, and human DNA samples, using even lower sample amounts than they would have needed with the RS II.

We are extremely excited about this latest chemistry improvement. With it, customers should be able to generate six to seven times more sequencing throughput per Sequel SMRT cell for their applications than they can with the RS II system. And it paves the way for expansion of a variety of applications on Sequel that have seen relatively little use on the RS II because of its sample requirements.

We showcased the improved performance during our workshop at the American Society of Human Genetics meeting in Vancouver two weeks ago that was enthusiastically received. Dr. Jonas Korlach, PacBio's CSO, presented results from both our internal and external beta testing.

Dr. Euan Ashley from Stanford University reported research data based on a tenfold coverage of a genome sample from an individual that was afflicted with a rare genetic disorder. The genetic lesion, a deletion in exon 1 of a key gene involved in tumor suppression, was clearly identified in the blinded analysis using the Sequel data but was not pinpointed in short-read-based sequencing experiments.

In mid-October, we announced that Sequel would be used in the next phase of the G10K and B10K projects. These programs were started to generate high-quality de novo assemblies of up to 20,000 vertebrate species. Participating institutions include scientists at Rockefeller University, University of California at Davis, UC San Diego, UC Santa Cruz, the Sanger Institute, Duke University, and BGI.

In August, we announced that HistoGenetics, a worldwide leader in HLA typing, won a large contract to process samples specifically with their PacBio systems. As an update, HistoGenetics presented at the annual American Society for Histocompatibility and Immunogenetics, or ASHI, meeting, in September, that they had already sequenced over 60,000 samples using SMRT Sequencing. They plan to sequence several thousand samples per week on their RS II systems for the next several years. They have also purchased a Sequel System for ongoing research.

On the publications front, we continue to see an accelerating pace of new publications featuring SMRT sequencing, with the total now over 1,800.

Two notable papers described de novo assembly of the genomes of individuals from Korea and China. In both cases, the authors detected more than 12,000 novel structural variants, as well as filled in several gaps in the GRCh38 NCBI reference genome.

Already, there are numerous similar programs to generate ethnicity-specific high-quality reference genomes that are in progress in many countries around the world.

Today, we announced that we have filed a complaint with the US International Trade Commission, the ITC, against Oxford Nanopore Technologies, Limited, ONT, and its affiliates. The complaint asserts that ONT's MinION and PromethION devices, and associated consumables, reagents, and software, as well as related commercialization activities by ONT and its affiliates, infringe one or more claims of US Patent No. 9404146, the 146 patent, owned by Pacific Biosciences.

The 146 patent, entitled: Compositions and Methods for Nucleic Acid Sequencing, was issued in August of this year and covers novel methods for sequencing single nucleic acid molecules using linked double-stranded nucleic acid template, providing improved sequencing accuracy.

We are asking the ITC to institute an investigation into the infringement of the 146 patent by ONT, and to issue an exclusion order permanently barring infringing ONT products from entry into the United States and preventing ONT from advertising and selling infringing products in the United States.

As we indicated in our press release earlier today, PacBio is a pioneer in the field of single-molecule long-read sequencing and we've invested hundreds of millions of dollars in research, development, and commercialization of this technology. This effort has spanned a diverse array of disciplines including nanofabrication, physics, organic chemistry, photonics, optics, molecular biology, engineering, signal processing, high-performance computing, and bioinformatics.

The contributions made by our many talented inventors have been recognized with hundreds of awarded patents relevant to long-read sequencing. We stand firm in our resolve to protect that investment and our leadership position in the field we have created.

I'll conclude my opening remarks with a brief update on our Roche partnership. We and Roche continue to prepare for their Sequel-based product launch. While Roche had initially targeted their launch for the latter part of this year, we believe that a launch date in early to mid-2017 for some of the targeted assays they are developing on the Sequel System is more likely.

That concludes my initial remarks. I'll now turn it over to Susan to provide more details on our financial results.

Thank you, Mike, and good afternoon, everyone. I will begin my remarks today with a financial overview of our third quarter that ended September 30, 2016. I will then provide details on our operating results for the quarter and year to date with a comparison to the third quarter and year-to-date 2015, respectively. I will conclude my remarks with a brief discussion of our balance sheet.

Starting with our third-quarter and year-to-date financial highlights, during the quarter, we recognized revenue of $25.1 million and incurred a net loss of $17.5 million. This brings our year-to-date total revenue to $65 million and our net loss to $55.3 million. We ended the quarter with $87.3 million in cash and investments.

Turning to revenue, total revenue for the quarter was $25.1 million, $11.2 million greater than the $13.9 million recognized in Q3 of 2015. Year-to-date total revenue in 2016 was $65 million, also higher than the $56.5 million recognized year to date in 2015.

In 2015, revenue was helped by the recognition of a $10 million Roche milestone achievement in Q2 of 2015. Excluding this milestone revenue, 2016 year-to-date revenue was 40% higher than the revenue in the first three quarters of 2015.

Breaking down the revenue, instrument revenue quarter over quarter was substantially higher, with $11.5 million recognized in Q3 2016 compared to $2.2 million recognized in Q3 of 2015. Year-to-date instrument revenue was $27.8 million, more than double the $13.5 million recognized during the same period last year.

Consumable revenue continues to be strong, increasing 22% to $6.5 million for the current quarter, up from $5.4 million reported during the third quarter of 2015. Year-to-date consumables revenue has increased 14% to $16.5 million in 2016 compared to $14.2 million in the first three quarters of 2015.

Service and other revenue increased 26% to $3.5 million in the quarter compared to $2.8 million in Q3 of 2015. Year-to-date service revenue was up 27% to $10.2 million from $8 million in 2015.

And finally, Q3 2016 revenue included $3.6 million of contractual revenue associated with the amortization of the $35 million upfront payment that we received from Roche in Q3 2013. This was consistent with the same amount recognized for the same period last year.

For the year, in 2016 we have recognized $10.8 million of contractual revenue compared to $20.8 million in 2015. Ben will provide more details on our contractual revenue later in the call.

With regards to gross profit and margins, we generated a gross profit of $12.6 million in Q3 of 2016, representing a gross margin of 50%. This was higher than the $6.6 million of gross profit and 47% gross margin recognized in Q3 of 2015.

Year-to-date gross profit was $32.8 million, representing a gross margin of 50% compared with a gross profit of $27 million and a gross margin of 47% for the same period in 2015.

It should be noted that year-to-date 2016 profit and margin were higher than 2015 even with the $10 million Roche development margin recognized in Q2 of 2015, with 100% margin. Excluding the $10 million Roche milestone revenue in Q2 of 2015, year-to-date gross profit increased by $15.8 million and margins increased to 50%, up from 37% in 2015.

This achieved increase is primarily a result of the higher revenue and margins recognized from sales of Sequel System.

Moving to operating expenses, operating expenses in the third quarter of 2016 totaled $29.4 million compared to $3.9 million in Q3 of 2015. As a reminder, we recognized a one-time $23 million gain associated with amendments to our facilities leases in Q3 of 2016. Excluding this one-time gain, our operating expenses in the third quarter of 2016 were $2.5 million higher than those in Q3 of 2015.

Year-to-date operating expenses increased to $86.2 million from $55.1 million in 2015; again, primarily as a result of the one-time gain described. Excluding the gain, operating expenses year to date in 2016 were $8.1 million higher than those incurred year to date in 2015.

Non-cash stock-based compensation included in operating expenses increased $1.6 million quarter over quarter and $4.2 million year-to-date 2016 over 2015.

Breaking down our operating expenses, R&D expenses in the quarter were $17.5 million, $1.3 million higher than the $16.2 million of expenses incurred in Q3 of 2015. Year-to-date R&D expenses were $51.4 million, a $5.7 million increase over the $45.7 million of expenses in 2015.

Expense increases in R&D were primarily related to compensation, non-cash stock compensation, and chip development expense. R&D expenses this quarter included $2.1 million of non-cash stock-based compensation expense, a $900,000 increase over the $1.2 million expense in Q3 of 2015. Year-to-date R&D non-cash stock-based compensation expense was $6.2 million in 2016, a $2.5 million increase over the $3.7 million expense incurred in 2015.

Sales, general, and administrative expenses in the quarter were $11.9 million compared to $10.7 million in Q3 of 2015. Year-to-date SG&A expenses increased $2.4 million to $34.8 million in 2016, up from $32.4 million in the first three quarters of 2015.

SG&A expenses in the third quarter of 2016 included $2.4 million of non-cash stock-based compensation expense, up $700,000 from the $1.7 million recognized in Q3 of 2015. Year-to-date 2016 SG&A non-cash stock-based compensation expense was $6.9 million, up $1.7 million from the $3.5 million recognized in 2015.

Finally, in Q3 we recorded $700,000 of net interest and other expense, primarily as a result of the incurred interest and derivative expenses associated with the debt we took on in Q1 of 2013.

Year to date, our interest and net other expenses have totaled $1.9 million.

Now turning to our balance sheet. As I mentioned at the beginning of my comments, our balance of cash and investments was $87.3 million at the end of the third quarter. This represents a $15.3 million decrease during the quarter.

Inventory balances increased $2.3 million in the quarter to $16.5 million from $14.2 million at the end of Q2. Accounts receivable increased in Q3 to $11.9 million from $10.4 million at the end of Q2.

This concludes my remarks on the financial results for the quarter. I would like to now turn the call over to Ben.

Thank you, Susan. I will be providing an updated forecast of our 2016 financial performance. First of all, as Mike mentioned earlier, we booked orders for 20 systems this past quarter. We're not providing a specific forecast for future bookings; however, based on our current pipeline, we expect our bookings to increase in the fourth quarter.

Now moving on to revenue. While our Q3 revenue was in line with our previous forecast, our instrument bookings in Q3 were less than our target, resulting in a lower backlog at the end of the quarter. Therefore, we may not achieve our previous revenue forecast for the year.

We currently estimate our product and service revenue to grow between 55% and 65% for the year. Previously, we had forecasted a 70% increase.

As we have mentioned in our previous calls this year, the quarterly amortization of contractual revenue is scheduled to drop to less than $100,000 in Q4 this year so that contractual revenue for the year is expected to be less than $11 million.

Therefore, in total we estimate our total revenue for the year to come in between $86 million and $90 million.

Now I'll comment on Q4. As we enter the fourth quarter, we want to remind you that in the fourth quarter of last year, we recognized a $20-million milestone associated with our Roche agreement. Since we also recorded amortization revenue of $3.6 million in Q4 of last year, we expect our contractual revenue in Q4 2016 to decrease by $23.5 million from Q4 of 2015.

Moving on to gross margin, we were pleased to again see the improvement in gross margin, up to 50% this past quarter compared with a margin of 47% in Q3 of last year. For the fourth quarter, we expect our gross margin percentage to decrease to the mid-40%s because of the drop in contractual revenue, which had a 100% margin associated with it.

For the year, we expect our average gross margin percentage to be a little less than 50%.

Our operating expense in Q3 came in as expected at $29.4 million. For the fourth quarter, we expect a modest sequential increase in operating expenses. For the year, we expect our total operating expense to grow by 10% to 12%. This comparison excludes the one-time $23 million gain we recognized in Q3 last year associated with the amendment of our property leases.

Our combined non-cash stock compensation expense and depreciation expense continues to be between $6 million and $7 million per quarter.

Regarding our interest expense, we continue to expect to report approximately $3 million for the year.

And finally, with regard to cash usage, for the first nine months of the year we consumed about $53 million in cash, excluding the cash we raised with the ATM. We are on pace to consume about $70 million for the year.

And with that, we will open the call to your questions.

(Operator Instructions) Bryan Brokmeier, Cantor Fitzgerald.

Hi, good afternoon. Mike, are you still placing RS IIs, and how much longer should we expect customers to be buying those?

We do place an occasional one. I can't really tell you what the long-term future is; we're not actively developing that for sales other than in special cases, some of which are non-sequencing uses, some of which are in sites -- HistoGenetics is one that we mentioned last quarter -- who had a workflow put in place based on the RS for a specific application, and we're just adding to their capacity. Our focus is on Sequel.

Okay. And could you elaborate a little bit more on the applications the lower sample input opens up that wasn't available with the RS II or the prior Sequel chemistry?

Well, it certainly helps in the context of looking at structural variant analysis, where people are interested in doing a large number of samples and they don't want to have to go through sort of a major effort to generate the DNA required to do the analysis there.

A lot of it can be in areas where for particular kind of samples you don't necessarily have that much DNA that you can get ahold of. And while that's not a problem in some areas, in a lot of areas it is.

An example would be in the case of the B10K and the G10K program, where a lot of those samples are from fresh animals and you get more DNA. Some of them are from very tiny animals where you can't get much DNA. And some of them are from deceased samples, where you have very little access to fresh DNA. So there are a whole series of things like that.

In areas of doing ISO-Seq analysis, we think it will reduce the amount of DNA required dramatically for a couple of reasons. And that's been a very potentially interesting area for us except in people's cases where they don't have enough DNA or can't generate enough cDNA from their RNA work to be able to use the technology.

So it's not that it's completely new applications, it's basically ones that were used on the fringe where there was a particularly compelling reason to be able to generate the long-read data but in which they couldn't really do a substantial number of samples with it.

And is that opening up more interest coming from biopharma? It may not have generated -- turned into revenues yet but are you starting to have more substantive conversations?

I think the area that is somewhat -- well, uniquely interests to them -- are both the structural variant area and in particular the ISO-Seq application for looking at transcript isoforms.

Okay. And then lastly, where's the high-throughput manufacturer in ramping up production and is the prototype manufacturer still producing at a higher rate than it had historically?

Well, as we pointed out last quarter, we would be in essence ramping down the lower-volume manufacturing output through this year with sort of stopping new receipts of wafers from them roughly near the end of the year while we're replacing their capacity and actually increasing it with the high-volume manufacturer. So it's one going down and one going up at a faster rate.

And that we would be continuing to sell chips from both suppliers through a good part of Q1 of next year, but by the end of that quarter we should be completely on the high-volume manufacturer as a supplier.

Okay, thanks a lot.

Bill Quirk, Piper Jaffray.

Great. Good afternoon, everyone. This is actually Alex Nowak filling in for Bill today. I just had a couple of questions on the orders and the backlog. So first one was Illumina yesterday reported that they're seeing a change on how US academia is doing their funding practices. I was just curious, do you think that had any impact on the order rate in the quarter, or do you think it all falls back on that there was a lack of Sequel data for the majority of the quarter as well as a limited SMRT cell supply?

I don't think that we saw any issue with funding processes per se in the US. I don't know -- they're in a somewhat different circumstance, particularly with some of their higher-priced offerings. From our perspective, it's the issues that I outlined in my comments. It's just a little bit of a delay in the summer getting people ramped up once we fully released the constraints on the chip supply, which in turn was holding them back from generating data that they could present at meetings. And we got through that process, but not really until the first meeting that came out, which was ASHG at the beginning of this month or the past month, now in October.

That's helpful and I actually want to expand on the ASHG conference; I know it was just two or three weeks ago but I was just wondering if you can give us any data point, if you're seeing an increase in conversations on Sequel after releasing the data at ASHG?

Yes, we saw a lot of increase in it as we were engaged in discussions with people at ASHG and subsequently to that with a broader customer base.

Okay, perfect. I'll jump back in the queue; I have some more questions.

David Westenberg, CL King.

Hi, guys; thanks for taking my question. The first one is just in the ordering patterns of customers in terms of SMRT cells. Specifically, are they ordering them in bunches in anticipation of projects, or is it more a steady stream? And do you anticipate a similar pattern with Sequel customers?

David, this is Ben. It's more a steady stream and there's a good reason for it. There's actually a shelf life to these things and so it doesn't help customers to stockpile those things that have a limited shelf life. So long as we can deliver their orders in a streamlined fashion, which we've been able to do, then there's no reason for them to order other than what they actually kind of need to run for the near term.

Got you; that's very helpful. And then, are you seeing an increased interest in customers focused on the microbiome projects? And how meaningful of a market do you think that's going to be for you guys?

Well, it's actually been a reasonable market for us for a while, and it should increase a little bit with the Sequel System. It's another one of those areas where sample amounts were somewhat restricted frequently. And so being able to operate with a much smaller amount of DNA on the Sequel System now should facilitate that.

We've had a lot of interest in both sort of whole-genome level, or attempted-whole genome level, microbiome studies as well as very focused ones that are going after things like 16S molecules. And there, being able to get a large number of counts ? or much larger than what we had with the RS -- per SMRT cell can make a difference as to how broadly the results can survey the microbiome population in a particular sample.

Got you. And then --

It's part of our portfolio of applications that we've established. It's working very well on the SMRT technology platforms. Like the others, with the Sequel System, it should grow.

Got you. And in terms of that sample input requirement, I have heard that customers with a 50-fold increase, it is promising. How much further do you think you could go in terms of input reductions?

Well, so the 50% input in reduction is at a certain level of size of DNA. And one of the aspects of the system that we're using, it has a much more dramatic impact on the amount of DNA required for really long-library fragments. And we've seen in certain cases much higher than a 50-fold increase. So I think that's a work in progress.

I don't know that we're trying to get down to the single-cell level for a DNA sequence with this particular chemistry, but we obviously have other approaches to being able to do things like that underway as well.

Got you, okay. And then, finally -- last one is on expectation management. How do you feel you're walking that line between explaining the potential of Sequel and not over-promising here?

Not so well since we just had to change our promise. So we're working on getting better at it.

Well, you're talking about in terms of performance or in terms of orders? I think in terms of performance, we --

Performance, yes.

Okay.

People have been getting results like we quoted here for a long time, as I pointed out, on things that were targeted-type sequencing applications, where you're doing relatively shorter pieces of DNA, maybe prepared by PCR for a targeted application because there, they weren't sample-limited. The PCR, the amplification by itself, generated as much DNA as you needed.

The problem we've had was in being able to allow them, with reasonable efforts and sample prep, to get enough DNA to routinely do that with long pieces of DNA, which is what most our de novo assembly customers would have required for any significant amount of work.

And as we solved that problem, then the issue with being able to get the yield of sequence and so forth with those kind of samples disappears. And so I think we feel pretty comfortable now that we've gotten through that hurdle and as people have tried it, who use it -- and we've now had it tested in a larger number of customers than even our beta program -- we feel pretty comfortable that that is a very solid improvement in the loading requirements.

And we had talked about that we would need to get to RS-level performance with the Sequel System by the end of this year, so this announcement and ability to ship this chemistry in October is a very nice place to put our customer base into.

One other thing I would add, David, is part of the strategy this year was to do a controlled ramp and to have a sort of high-touch point with the smaller number of customers, and that's because different customers, depending on their applications, do experience different things with the platform.

And so when you mention walking a fine line in terms of promising things, I think we are trying to be careful to be specific about which applications are doing what with Sequel, and which applications, as Mike has just highlighted, are now certainly enabled in a much greater fashion with the latest performance improvement.

Got you. Thank you; I'll hop back in queue.

Joe Munda, First Analysis.

Great, thanks for taking the questions. Mike, in the past you were able to break out for us somewhat new customers versus existing on the Sequel. Any chance at getting that this quarter?

You mean for the new orders, Joe?

Correct.

Yes. I mean, I would just say that generally speaking what's naturally happening is more and more the customer orders are coming in from new customers because the population of existing customers to draw from is, I guess, dwindling a bit. So pretty much each quarter what we've seen is a really healthy mix of new customers, and I think that trend will continue.

Okay. And any sense of what the shipments were to Roche in the quarter? I mean, handful? Matching up to last two quarters? Is that about the ballpark?

Yes, Joe, we're not breaking out specifically the sales to Roche versus non-Roche; we don't want to get into a habit of doing that level of scrutiny on, I guess, the shipments that we have. Generally speaking, every quarter we've had a small number of shipments to Roche and this past quarter was no different.

Part of that's in response to Roche's requirements, but that would be our distributor and how they ship them and how they release any numbers based on that performance is up to them, not us. So as we go forward, we will be pretty careful not to break out our particular business with Roche.

Okay. And Mike, I attended ASH, and I attended the seminar that Dr. Korlach presided over. Pretty impressive. Our conversations with the attendees was very positive. So based on the guidance that you guys are putting out there, and kind of linking that with the response that we saw, I mean, can you give us a sense of how much of a gating factor the chemistry may have been up until this point, and I guess give us a little bit more color on what exactly it means going forward as far as utilization and attracting new customers? I know that's a lot, but the feedback we had from attendees was quite positive, so I'm trying to make a distinction here of what's really going on.

Well, I think the feedback that you got is the same feedback that we got. In fact, it even started for us earlier than that because we had the beta program going earlier than, obviously, the results that were presented at ASHG.

As I said in my comments, we were extremely encouraged by the response and excited by the performance gain that we?ve got in that loading issue, not just because it will enable people to do what they were expecting to do, particularly in the de novo assembly and the structural variant world, where you're dealing with whole genomes and large fragment libraries, but because of the ability to broaden our penetration in some of the other applications as well.

So it hasn?t been that long -- two weeks since ASHG -- but we've seen a renewed interest and an interest from people that maybe were exposed to the technology and where it actually is for the first time at that meeting. And that's starting to spread to other circles as well.

So time will tell how fast that translates into an orders ramp, but we certainly are optimistic about that happening.

Okay, that's quite helpful. And then I guess my last question as it relates to -- another caller brought up Illumina. On their call they had mentioned 15% of high-throughput sequencing devoted to whole genome. I guess, taking that statistic and extrapolating it to PacBio, can we be more positive now that the need for whole-genome sequencing and what you guys are doing is being proven in the marketplaces, based on Illumina's comments?

Well, the definition of whole-genome sequencing is in the eye of the beholder, or the supplier, as the case may be. I think Illumina has seen that interest for a long time in the context of their short-read whole-genome systems, particularly with their X Ten platform, which is geared almost exclusively to that.

From our perspective, that does a great job at a fairly low price of handling a large number of samples. As was, I think, shown again at the presentations at ASHG at our workshop, it still leaves a lot unanswered about genetic variation from one individual to another.

And so I think what's happening is an increasing recognition that you need more and more of both attacks on looking at structural variation as well as single-nucleotide variant analysis, the latter being something that the short-read technologies are very good at, not so good all the time at looking at larger structural variants. And both have a big implication for human disease.

So I think there's a recognition that's certainly been growing for a while of the importance of long-read sequencing, particularly in regard to structural variant analysis. And as our price per sample continues to go down, then more and more people can do larger projects alongside, maybe, their short-read sort of single-nucleotide analysis.

And then we've also seen -- and that's the announcement with the G10K and B10K -- that in plant and animal, where you don't have a reference, that this is an area where the PacBio SMRT long-read technology is really continuing to penetrate and grow because it's giving answers that didn't exist before.

Okay, thank you; I appreciate the comments. I'll hop back in the queue.

(Operator Instructions) Amanda Murphy, William Blair.

Hi, good afternoon. I guess this is another question on the pipeline and whatnot. I think one of the questions we get asked a lot is in terms of the long-read market -- is it really just a niche market? And I think having the orders decelerate in the quarter might provide some level of credence to that, I guess.

So what I think would be helpful is if you were to somehow -- and I know you don't want to put specific numbers -- but is there any way you can help us quantify your pipeline just now, coming out of ASHG, what that might look like vis-a-vis earlier this year, just to get a sense of what the -- whether we are at an inflection point or not currently?

I'll try first and then maybe Mike wants to try. We hesitate to give exact forecasts, but we purposely are saying that we expect fourth quarter to be greater than third quarter. As Mike mentioned before -- I'll use a different term, though -- I would say that there's positive momentum in sales now that is higher than it was before we had this latest release of the improvement due to performance.

That gives us confidence that we are definitely going to be seeing growth in the orders. And we're not giving yet a forecast that's specific for 2017, but we certainly expect the growth in the orders to continue through 2017. Again, I hesitate to be specific on the numbers, but I certainly don't think that -- while I hear what you're saying; it's a lower number in Q3 than let's say in the previous quarters -- we certainly do not think that that's a trend that will continue.

It wasn't that our pipeline in Q3 was going down; it actually was increasing. It's just that people were waiting until they saw results from the early labs, particularly in the areas that they were most -- the application side -- that they were most interested in. And the minute that we could show data, and our customers could show data and become enthusiastic reference sites, that enthusiasm can start to turn into sales. And I think we're beginning to see that already, even a couple of weeks after the meeting.

The way I would characterize it is -- what we said is that we expected to be able to have more orders in the second half of the year than the first half, and we'll be close to that. But the other way to look at it is that we kind of were probably slower in that ramp-up turnaround by maybe a quarter just because it took a little bit longer to get the message out to people.

And particularly with the breakthrough that we've had on the loading requirements, that's attracted an even different set of interest from customers in the Sequel platform. So I think we feel pretty good about the backlog in terms of potential orders continuing to go up.

Okay, fair enough. And then, I know this is a difficult question but we get asked it all the time, so I figured I'd ask you, too. Just thinking about the market opportunities three to five years from now, I think people for around 10% of all applications are long read. And I realize that obviously long read doesn't have to just be long read; it could be used for other things, too.

Is there any way to kind of put some quantification around what you guys think the market opportunity is fast forward once you're kind of in the scale mode of the Sequel?

Well, at least 10%'s better than the 2% that other people thought it was maybe a year and a half, two years ago. But it really is an issue of price per sample. And as we continue to push the price per sample on our system down faster than it can realistically be expected out of the short reads, it isn't that people don't think that they'd be better off using long reads for everything as long as it didn't cost them too much to be able to run their sample cohorts. And so as that issue dissipates, I would argue that it's an almost unlimited potential of the market from a percentage to go after.

That said, there are certain tests where you don't even utilize the full reading capability of short-read sequencing. NIPT is an example of that, where you only need 25, 30 base pair reads to be able to map fragments onto a reference genome. Because you're not interested in the sequencing per se, you're interested in counting and in relative counting of one chromosome versus another and maybe seeing one event versus another.

But for most of the bulk of sequencing applications, most people would prefer to do it with long reads as long as they can afford to do it. And so it's up to us to get that cost of our system usage down into the same runs as the short-read technologies. We do that, everything is on the table.

Do you guys still have that same target you laid out initially -- I think you said $1,500 to $3,000 per genome in 2017, 2018? How has that changed based on how this year has progressed thus far, or has it?

Well, the way that we continue to lower the costs are simply to increase the multiplexing capacity of a chip to go to a higher-density SMRT cell than we have now, and that program, from a development perspective, is underway. I'm not prepared to give an update; we'll maybe talk about this a little bit more in our beginning-of-the-year call, when we do forecasting. But we do not see a technical hurdle to being able to do that at this point. And the development program with our partners is well underway.

It's limited by how fast it takes to get the chip developed and then manufactured, but we've at least gone through the process with both our development partner now and our high-volume manufacturer on the general process, and it's just a matter of using -- the R&D guys would kill me if I said this to them directly -- but just a matter of engineering it at this point. It?s using a newer-generation sensors and slight tweaks to the arrangement of the ZMWs to match up with that.

Amanda, I would just say that I just mentioned earlier in this call that now multiple customers are getting 5 to 7 gigabases on a SMRT cell. I think you're going to have to see what that actually does in terms of the impact on the demand. Because that dramatically brings down the cost to a level that a lot of customers have been kind of looking for.

So again, going back to what's going to lead to a certain momentum and adoption, we certainly think that that's, in itself, going to lead to a decent amount.

And I would point out that this isn't the last upgrade on the performance that we will have, even on the current chip. That we still are looking at a change in the fundamental sequencing chemistry reagents near the end of this year that will boost the read lengths still more and concurrently increasing the throughput.

So we're in a fairly aggressive performance increase mode at this point with our developments. And we think that the software platform is pretty stable, allows for additions of new protocols, like the loading protocol that we?ve just introduced, as well as a switch-over of the actual sequencing chemistry.

So we're not going to wait for the higher-multiplex chip; that has its own development timeline. But we expect to increase the performance of the current 1 million ZMW chip substantially as we're still developing the higher one, which will take advantage of whatever the chemistry and sample-prep protocols are at the time that the higher chip is available.

Got it. And I just had one last one on Roche. I thank you for the comments on the timeline; that was helpful. But I just wanted to clarify because I think there's some confusion out there about this point in that obviously Roche is your distributor in the clinical market. But can you just talk to your abilities to service the clinical market? So in other words, are you precluded from shipping a research Sequel to a lab that has clinical functionality?

Well, precluding is not necessarily the right word. Our agreement with Roche is that as they get to the point where they're releasing the system, they will be our distributor into the diagnostic marketplace. Clinical, in a research mode, has different use cases depending what the customer's doing.

That said, in the -- if someone wants to operate with an RS, we have no restrictions and Roche has no distribution rights for that platform. And as, I think, in response to an earlier question about whether we continue to ship those, we do, and one of those cases that we?ve talked about, as I pointed out, was HistoGenetics, which is nominally a Roche -- diagnostic-related field; it's kind of a special case. But in that case, HistoGenetics chose to expand their capability with the RS, which met very carefully and very well their requirements there in terms of throughput and workflow. And they're now our single-biggest customer by quite a bit.

Right. And Mike's got another example on that with the HistoGenetics. They may be called a clinical lab, but they bought a Sequel for research. So it depends on your definition, and that we go into detail on with Roche privately, on what is the machine being used for and what is the final intention? So you will see potentially mixed sites out there just by the nature of what that site is doing at different points in the site.

Got it. Okay, thanks very much.

Bill Quirk, Piper Jaffray.

Thanks, this is Alex Nowak again, for Bill. Just a quick follow-up question. It's a two-part on expense forecasting for 2017. First, can you say how much that you think the Oxford Nanopore lawsuit will add to G&A expense for 2017?

And then, the second piece is with Sequel development slowing down, or likely slowing down here with the instrument chemistry kind of stable, obviously precluding the development of the higher-density chip, do you think you'll see some R&D expense roll off for 2017?

Well, let me answer the latter part of that question. I thought I just answered to Amanda that we still have a very aggressive development program on both the current Sequel chip in terms of the chemistries associated with that and the software that's associated with the applications, while at the same time we are working on developing a second generation of the chip, which has a higher ZMW count.

So it's not that we're slowing down our efforts; maybe we get more efficient relative to the amount of money that's required to get one thing done. But we still have a very aggressive development program to improve the performance because we do want to get at that much larger portion of the market that is much more cost-constrained in their projects.

I'll just say that we're still working on our 2017 plan so we're not ready to comment on the specifics of what our spend is going to be for 2017. If there's any indication -- at least in the fourth quarter I mentioned that there's going to be modest spending increase from Q3 to Q4. It takes everything into account, including any sort of legal expenses. So hopefully that's --

In terms of the legal costs, one of the things that I think is important to point out, that we chose to do this initial filing through the ITC. And one of the reasons is that the ITC process, as long as you meet the requirements for what their purview is, is a much faster process than is typical in US District Court with a patent infringement case. And what it means is that you have a much more defined timeline and hence expense associated with it. It is not nearly as open-ended as a traditional District Court patent case can be.

We're not prepared to give you what we think the estimate of that is; maybe we'll talk about it a little bit more in looking at 2017, where most of those expenses would occur. But it is the most efficient process for dealing with patents, particularly between a US company and a non-US company who's importing a potentially-infringing product into the US.

I want to add one more thing. You are asking about 2017. And this backs up a little bit to some of the questions that Amanda had. In terms of the bookings, we are planning to again report the bookings at the end of Q4. But starting in 2017, we don't see a need to report bookings anymore, and I think now that the sort of shipments have caught up to the bookings, we're just going to rely on the reporting of revenues in 2017. Just wanted to give you a heads up on that.

Okay, that's helpful. Thank you very much.

I think that's probably all the time we have left for this call. So I think we're going to go to closing.

Thank you ladies and gentlemen, that concludes our question-and-answer session for today. I would like to turn the call back over to Mr. Mike Hunkapiller for closing remarks.

Thank you. In closing, I'd like to reiterate our excitement around our latest Sequel chemistry improvement. As I said before, with this release customers should be able to generate six to seven times more sequencing per SMRT cell than they could with the RS II system on a broader range of applications, and we look forward to seeing our customers expand the variety of applications that they perform on a Sequel System as a result.

As always, we remain steadfast in our commitment to bringing the unique advantages of our SMRT technology and products to our customers and the scientific community in general. We believe that SMRT sequencing provides the industry's most complete and accurate picture of genomes due to its superior performance in sequencing accuracy, uniformity of coverage, extremely-long read lengths, and ability to characterize DNA-based modifications.

Furthermore, by providing scientists with an ability to obtain a comprehensive set of sequence information within a single experiment, SMRT sequencing is often the lowest-cost and only research tool available to meet their needs.

Thank you for joining us and we look forward to talking again in three-months time.

Ladies and gentlemen, thank you for your participation in today's conference. This concludes the program and you may now disconnect. Everyone have a great day.