Pacific Biosciences of California, Inc. (PACB) Q4 2013 Earnings Report, Transcript and Summary

Good day, ladies and gentlemen. Thank you for standing by, and welcome to the Pacific Biosciences Fourth Quarter Fiscal Year 2013 Earnings Conference Call. [Operator Instructions] I would now like to turn the conference to our host, Ms. Trevin Rard. Ma'am, you may begin.

Thank you. Good afternoon, and welcome to the Pacific Biosciences Fourth Quarter and Fiscal Year 2013 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 Form 8-K available on Securities and Exchange Commission website at www.sec.gov. With me today are Mike Hunkapiller, our Chairman and Chief Executive Officer; Susan Barnes, our Chief Financial Officer; and Ben Gong, our Vice President of Finance and Treasurer. Before we begin, I'd like to remind you that on today's call, we will be making forward-looking statements, including plans and expectations relating to our financial projections and products that are subject to assumptions, risks and uncertainties and may differ materially from actual results. These risks and uncertainties are more fully described in our Securities and Exchange Commission filings, including our most recently filed quarterly 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'll now turn the call over to Mike.

Thanks, Trevin. Good afternoon, and thank you for joining us today. We are pleased with our fourth quarter results and the progress we've made throughout 2013 on driving our overall business. Highlights of our fourth quarter and full year results are as follows: We booked orders for 9 PacBio RS II systems in the fourth quarter, bringing our total bookings for the year to 25. This represents more than double the bookings we had in 2012. We installed 5 systems during the quarter, which brings our installed base up to 88 and our backlog of systems to 13 as of the end of December. Consumable revenue for the fourth quarter was $2.6 million, which was double that of Q4 2012 at $1.3 million and up 25% sequentially from our third quarter consumable revenue of $2.1 million, showing strong utilization in our installed base. You may recall during our last quarter's call that we were cautious about our fourth quarter consumable sales because we started out a bit slow due to the U.S. government shutdown. As it turned out, our business with U.S. government customers recovered, and we had strong growth among numerous other customers in the quarter. For the year, consumable revenues grew 84%, stemming from an increase in our installed base of about 25% and more significantly, by a greater than 50% increase in the average revenue pull-through per system. We are averaging well over $100,000 in consumable revenue per year per install system. Total revenue for the fourth quarter was $9.1 million, up 55% from the $5.9 million in Q4 2012. For the year, total revenue grew 8% from $26 million in 2012 to $28 million in 2013. Our development team did a terrific job delivering new products in 2013. Over the course of the year, we introduced a series of product enhancements that quadrupled the throughput of our system and greatly increased the read lengths achieved by our customers. With our latest chemistry release, customers have reported average read lengths in excess of 8,500 bases, and so their longest reads have exceeded 40,000 bases. In addition, we delivered significant enhancements to our software analysis toolkit, enabling customers to achieve the highest consensus accuracy with PacBio compared to all other sequencing technologies, including Sanger sequencing. This past year, PacBio became largely recognized as the gold standard for sequencing bacterial genomes and characterizing pathogenic strains of microbial organisms. Previously, we have described the 100K Pathogen Genome Project established by UC Davis, the FDA and other collaborators to create a database of food-borne disease-causing microbes, and that project has been progressing. In similar fashion, during the fourth quarter, Public Health England, the National Collection of Type Cultures and the Wellcome Trust Sanger Center announced a collaboration that will use PacBio's SMRT sequencing to fully sequence and finish a collection of 3,000 bacterial genome strains, which they expect to use as a reference for basic and clinical microbiology. That project is headed by Dr. Julian Parkhill, the head of Pathogen Genomics at the Wellcome Trust Sanger Institute and a highly regarded leader in the area of bacterial genomics. In the latter half of the year, we entered into a collaboration with Roche Diagnostics, the world leader in in-vitro diagnostics. In the short term, this agreement provided us with cash to further invest in developing our products in key applications. In the medium and long term, we believe this will enable us to capture a significant portion of the molecular diagnostics market as the provider of sequencing products to Roche. We've gotten off to a good start with Roche, working toward our common goal of creating products for the diagnostics market using our SMRT technology. Roche's sequencing unit is conveniently based in the San Francisco Bay Area, which makes it easy to have face-to-face communications. Our intent is to work on regulatory clearances in parallel with product development activities in order to be in position to sell products for this market as soon as possible. As we have mentioned previously, we believe initial regulatory clearances for our products could take 2 to 3 years. Finally, we ended the year with $113 million in cash, which is up $12 million from our cash balance at the end of 2012. Now I would like to provide highlights from 2 industry conferences we attended since our previous earnings call. The first is the American Society of Human Genetics conference that was held in Boston during late October. At the ASHG, we hosted a workshop featuring multiple presentations on structural variation studies with SMRT sequencing. As evidenced by the packed lecture room, the talks generated lots of interest and they did not disappoint. Dr. Evan Eichler of the University of Washington presented data from back sequencing in regions of the human genome that have previously been inaccessible to study, effectively with first- and second-generation sequencing tools, but which he is now able to study with SMRT sequencing. In his talk, Dr. Eichler expressed his excitement in having access to study the complex regions of human genomes at low cost and to recover areas that most genome studies have skipped over. He presented to a very full audience. We're clearly impressed with the data, and since then, we have seen momentum building in our business around human genome sequencing. Last month, we attended the annual Plant and Animal Genome conference that was held in San Diego. There were 15 program presentations and 19 posters featuring PacBio biotechnology this year, which is roughly double the number from last year. The presentations were wide-ranging from sequencing cash crops, such as spinach, lettuce and wheat, to the advanced study of model plant organisms such as Arabidopsis. There seemed to be an overriding theme to the presentations, which was the SMRT sequencing has been enabling researchers to reveal new areas of genomes, often discovering new genes and uncovering structural variations that have previously gone unnoticed because of the shortcomings in sequencing the short-read technologies. For example, in the presentation given by researchers from Ghent University, it describes how to use PacBio to sequence isoforms of the Spruce genome which they thought have been fairly well characterized by high coverage Illumina sequencing. They were surprised to find 2,000 unique protein-coding genes from the PacBio analysis. In another study, researchers at the Salk Institute described their findings from sequencing strains at the model organism, Arabidopsis, with PacBio. When compared against short-read sequence data, the PacBio data revealed that 1/3 of the single nucleotide polymorphisms or snips present in the strains were missing from the short-read data. Studies such as these are raising interest in the agricultural area because of the significant ramifications of missing large chunks of data when trying to determine the markers for important crop traits such as those for disease and drought resistance. Concurrent with the PAG conference, we worked with Dr. Casey Bergman from the University of Manchester, Dr. Susan Celniker and Roger Hoskins at Lawrence Berkeley National Laboratory and Dr. Sergey Koren and Adam Phillippy at the University of Maryland to post new data sets on the assembly of the model organism Drosophila, a fruit fly, using SMRT sequencing. Model organisms can be the best, fastest and most effective way to advance science, especially when human experimentation may not be feasible, and Drosophila has been used to study genetics for decades. This new data set for Drosophila using PacBio technology represents a dramatic advancement for researchers, significantly improving the assembly of this genome by filling in regions of sequence missing in reference genome. For example, chromosome 3L was reduced from 22 contigs or fragments down to just a single contig, meaning that this chromosome are almost fully assembled with no gaps using the PacBio data alone. This level of completeness in the de novo assembly is unprecedented in the metazoan genome. To put this in perspective, the reference genome for Drosophila, known as Release 5, has been over 15 years in the making. With experts utilizing Sanger sequencing, short-read sequencing, RNA transcript sequencing, back clones and numerous labor-intensive techniques to achieve what is considered a high-quality genome, the PacBio technology, a few researchers needed just 6 weeks from collecting the sample to completing a considerably higher-quality assembly. Dr. Celniker, who is regarded by many as the foremost expert of the Drosophila genome, commented that the PacBio assembly was a huge improvement over the reference, and she was particularly impressed with PacBio's ability to piece together regions of the genome that have thus far been a mystery such as the Y chromosome. This data set has also been made publicly available, and we believe this will lead to a flurry of new research in this important area. As a footnote, about 2 weeks ago, a group of researchers posted their latest results of sequencing and assembling Drosophila using so-called synthetic long-read technology as an adjunct to standard short-read methods. Their results suggest some improvement with this approach relative to using standard short-reads alone. However, they were nowhere near those obtained with SMRT sequencing. The PacBio assembly left far fewer gaps at 128 contigs for SMRT sequencing versus more than 5,000 with the synthetic long-reads. N50, which measures the length of the contig for which half of the total assembly is contained in contigs of that size are larger, was remarkably better in the PacBio assembly. 15 megabases versus 0.1 megabase, a 150-fold improvement. Now shifting to publications. The rate of customer publications with SMRT technology continues to increase. On a cumulative basis, we've been over 100 publications, up from about 20 a year ago. I'd like to highlight some recent publications focused on an emerging application for us: isoform sequencing for transcriptome analysis. The first paper I'll highlight appeared in the Proceedings of the National Academy of Sciences, or PNAS, entitled Characterization of the Human Embryonic Stem Cell's Transcriptome by Hybrid Sequencing, written by a group at Stanford. In the paper, the authors provide background by describing a problem they see with previous and ongoing RNA-Seq studies, which is the inability to capture full-length mRNA isoforms by using reads of just a few hundred-base pairs. The authors write, "We are still far from achieving the original goals of RNA-Seq analysis, namely the de novo discovery of genes, the assembly of gene isoforms, and the accurate estimation of transcript abundance at the gene or isoform level." Isoform detection or prediction with short-reads is even more difficult when the full set of possible isoforms is not known going into the project. The authors then described their own RNA-Seq experiment utilizing PacBio data to examine the human embryonic stem cells which were thought to be well characterized in numerous previous studies. Surprisingly, over 1/3 of the RNA transcripts the researchers detected were novel isoforms that were previously unidentified or missing from all previous studies. The authors concluded that these results suggest that gene identification, even in well-characterized cell lines and tissues, is far from complete. Another paper on this topic was published recently in Nature Methods, entitled Assessment of Transcript Reconstruction Methods for RNA-Seq, written by researchers from the European Molecular Biology Laboratory. In the paper, the authors described that "the complexity of higher eukaryotic genomes imposes severe limitations on transcript recall and splice product discrimination that are likely to remain limiting factors for the analysis of current-generation RNA-seq data." To address these limitations, the authors conclude that "ultimately, the evolution of RNA-seq will move toward single-pass determination of intact transcripts. Third-generation instruments will realize that potential and inspire new computing approaches to meet the next wave of innovation in transcriptome analysis." Now turning to our product roadmap. This is a good lead-in to describing the papers, our product and technology roadmap for this coming year. With growing interest in using SMRT sequencing for the analysis of full length transcripts and alternative splicing isoforms, we plan on releasing software in the first half of this year for streamlining these experiments. We call this application IsoSeq, which is short for isoform sequencing. Additional software enhancements planned for this year include enhanced detection of viral minor-variants in genome examples, improvements to long-amplicon haplotype analysis to benefit applications such as HLA analysis and continue the advancements in assembly algorithms for applications such as diploid de novo assembly. In the area of sample preparation, we will provide new protocols and reagent kits for automated library preparation methods on liquid-handling robots such as Agilent's Bravo and PerkinElmer's Sciclone workstations. We are also continuing to improve methods to better ensure the integrity of long DNA inserts, so users can take better advantage of the PacBio reads that are 20 kilobases and longer. With regard to read length, I mentioned earlier that customers have reported average read lengths in excess of 8,500 bases with our P5-C3 chemistry. We are continually driving improvements in read length through a combination of sample prep improvement, software enhancements and new chemistries. Our goal over the next 2 years is to increase the average read length to 20,000 bases. Keep in mind that the longest reads reported today are already in excess of 40,000 bases, and we should obviously expect that to increase as well. Finally, we plan on continuing our rapid pace of increasing the throughput capabilities of the PacBio system. In each of the past 2 years, we have increased throughput of the PacBio RS system by a factor of 4. Our target for this year is to once again increase it by a factor of 4. We have several ways to accomplish this. First, as I just mentioned, we can increase the average read length through continued improvements in chemistry. Second, we can increase the number of active wells in the SMRT cell. Currently, about 1/3 of the 150,000 wells on a SMRT cell are generating useful data because the current loading of enzymes in the well stems for process which follows a plus-on distribution that limits the maximum number of singly loaded wells to about 55,000. We are working on techniques for improving the loading efficiency of the SMRT cell beyond this distribution. Lastly, as I mentioned earlier, we are working on methods for improving the quality of DNA in the sample preparation process. The input quality of DNA has a direct effect on sequencing read length and throughput. None of these throughput improvements should require hardware changes to the PacBio RS II systems. Now I'd like to provide a few comments on the markets and applications we're addressing. Recent publications and customer presentations at industry conferences like the ASHG, PAG and AGBT have been raising awareness of the high value we bring to a variety of applications for SMRT sequencing. At last month's JPMorgan Healthcare Conference, we describe how we initially focused on sequencing DNA from isolated bacteria and demonstrated the high value of SMRT sequencing in that area. Over the past year and even more so over the past 6 months, we have demonstrated the value the SMRT sequencing brings to in areas such as metagenomics for studying the human microbio and the value it brings to direct human sequencing for studying portions of the human genome that have proven to be unsequenceable by other methods. We are still very early in driving on these markets, but we're seeing signs it's making a difference in our sales pipeline. Finally, I would like to mention that we are looking forward to hearing from many of our customers at next week's AGBT meeting in Marco Island. There are 9 schedule podium presentations featuring PacBio at this year's conference, up from 5 last year, and 29 posters this year compared to 19 last year. We will also be hosting a workshop on Friday at the conference where we'll be sharing some of our latest application developments, including IsoSeq, PacBio's solution for analyzing full-length transcripts. That concludes my remarks, and I'll turn the call over to Susan.

Thank you, Mike, and good afternoon, everyone. I will begin my remarks today with financial overview of our fourth quarter that ended December 31, 2013. I will then provide details on our operating results for the quarter with a sequential comparison to third quarter of 2013, a year-over-year comparison to the fourth quarter of 2012 and a full year 2013 to 2012 comparison. I will conclude my remarks with a brief discussion of our balance sheet. Starting with our fourth quarter financial highlights. During the fourth quarter, we recognized revenue of $9.1 million, incurred a net loss of $17.2 million. We ended the quarter with $113 million in cash and investments, a decrease of $14 million from the balance at the end of Q3. Breaking down our revenue. Total revenue from the quarter was $9.1 million, up $1.7 million from the $7.4 million of revenue realized in Q3 and up $3.2 million from the $5.9 million recognized in Q4 of 2012. For the year, revenue in 2013 was $28.2 million, an increase of $2.2 million from 2012. For instrument revenue, in Q4 of 2013, we recognized $3.2 million on 5 instruments installed compared to $3.7 million on 6 instruments in Q4 and $3 million from 5 instruments recognized in Q4 of 2012. For the year, instrument revenue was $11.5 million in 2013, a reduction from the $15.5 million recognized for instruments in 2012. As a reminder, higher instrument revenue 2012 was a result of working off an early commercial backlog of 16 systems coming into 2012 compared to 5 systems in backlog coming in through 2013. Consistent with the momentum we've been describing in our last earnings calls, our backlog coming into 2014 has grown to 13 systems. Consumable revenue was again strong this quarter. Consumable revenue for the quarter totaled $2.6 million, up 25% from the $2.1 million recognized in Q3 and up 97% over the $1.3 million recognized in Q4 of 2012. For the full year of 2013, consumable revenue was $8.5 million, an increase of $3.9 million or 84% over revenue recognized in 2012. Service and other revenue is primarily composed of service and grant revenue and with $1.6 million in both Q4 and Q3 of 2013. For the full year, our service and other revenue was $6.4 million in 2013 compared to $5.9 million in 2012. However, to understand the growth in this area, it's important for me to break out the service revenue. In Q4 2013, service revenue was $1.6 million, an increase of 9% over the $1.5 million recognized in Q3, an increase of 24% over the $1.3 million recognized in Q4 2012. Year-over-year, service revenue increased 22% to $5.8 million for the full year 2013 versus $4.7 million recognized in 2012. And finally, this quarter, we recognized $1.7 million of revenue associated with the Roche agreement that we announced last quarter. Gross profit was $3.2 million in Q4, representing a gross margin of 34.5%. This included the $1.7 million of profit related to the Roche agreement as that revenue has no associated cost of goods sold. The gross profit in Q4 is $2 million higher than the $1.2 million of gross profit recognized in Q3 and $2.6 million higher than the $600,000 we recorded in Q4 of 2012. For the year, gross profit was $6.4 million, a $5.5 million increase over the $900,000 recognized in 2012. These significant improvements are not only driven by the Roche contribution, they also include the benefits of strong growth in higher-margin consumable sales and continually improving service margins. Our increase in service margins is a direct benefit of our improvements in system reliability. Uptime for our system are now in excess of 97%. Moving to operating expenses. Operating expenses in the fourth quarter totaled $20.2 million compared to $21.2 million for the previous quarter and $22.3 million incurred in Q4 of 2012. For the full year of 2013, operating expenses totaled $84 million, a reduction of $11.3 million from the $95.3 million expense recorded in 2012. Breaking down our operating expenses. R&D operating expenses in the quarter were $11.1 million, up $700,000 from the $10.4 million recognized in Q3 of 2013 and were $600,000 below $11.7 million of expense in Q4 of 2012. Expense variances reflects a normal expense fluctuation associated with product development. For the full year 2013, R&D expenses were $45.2 million, a $2.5 million decrease from 2012 levels. Main drivers from reduction -- related to reductions of supporting infrastructure costs such as those associated with IT expenses and asset depreciation. R&D expenses this quarter included $1 million of noncash stock-based compensation expense. Sales, general and administration expenses for the quarter were $9.1 million, $1.7 million lower than the $10.8 million incurred in Q3. This decrease was primarily a result of $2 million of expense recognized in Q3 associated with the Roche agreement. Year-over-year, sales, general and administrative expenses decreased by $1.6 million from $10.7 million in Q4 2012. And for the full year, SG&A expenses were $38.7 million, a $9 million decrease from the $47.7 million recognized in 2012. The expense reductions realized year-over-year reflect both lower legal expenses in 2013, as we incurred settlement expenses of 2 patent disputes and a class action lawsuit in 2012, as well as continued operation of efficiencies realized throughout the organization. Ben will provide further guidance on our ongoing expense rate later in the call. SG&A expense for the fourth quarter includes $1.3 million of noncash stock-based compensation expense. Also, in the area of other income and expense, this quarter, we reported approximately $200,000 of net other expense, primarily weighted to the interest and derivative expense associated with our debt, which we took on in Q1 of this year. This is a $300,000 reduction from the net expense realized in Q3, as the reevaluation of the derivative associated with the debt financing can cause fluctuations in recognized quarterly expense. For the year of 2013, we incurred $1.8 million of net other expenses versus $100,000 of net expense recognized in 2012. Now turning to our balance sheet. Cash and investments totaled $113 million at the end of the fourth quarter, down $14 million from the previous quarter. Cash used in the quarter was $14.4 million, reflecting our fourth quarter net loss of $17.2 million with $4 million in noncash expenses from stock-based compensation expense and depreciation, offset with the end-of-year reductions to our liability balances. Among other accounts, accounts receivable decreased $1.1 million from $3.8 million at the end of Q3 to $2.7 million at the end of Q4. This quarter variance is due to the normal fluctuations associated with the instrument installed and collection timing. And inventory balances remain relatively flat, increasing $200,000 from $9.8 million at the end of Q3 to $10 million as of December 31, 2013. This concludes my remarks for the financial results of the quarter. I would now like to turn the call over to Ben.

Thank you, Susan. I will be providing guidance on our near-term and 2014 financial performance. Starting with revenue, we exited the year 2013 with a solid quarter of instrument bookings and continuing growth in system utilization which should drive revenue growth in 2014. We are forecasting significant increases in all revenue categories, resulting in at least 55% growth in total revenue for the year. This includes approximately $6.8 million in annual revenue, recognized ratably at $1.7 million per quarter, in connection with our agreement with Roche. If we exclude revenue from our Roche agreement from both 2013 and 2014, we expect total revenue to grow at least 40% from 2013 to 2014. Looking at revenue in the near term, having seen a significant jump in consumable revenues in Q4, we expect to see our consumable shipments flatten out somewhat in Q1. On the other hand, we started Q1 with 13 systems in backlog, which is up from the 9 that we had in backlog at the beginning of Q4. We expect to install at least 7 of the 13 backlog systems in Q1 compared with the 5 systems we installed in Q4. As a result, we expect total revenues to increased sequentially from Q4 2013 to Q1 2014. Moving on to gross margin. We saw our gross margin jump from the mid-teens for most of 2013, up to almost 35% in Q4 as a result of the $1.7 million in Roche revenue recognized as a 100% margin and a relatively high mix of consumable revenue, which has a higher margin than our instrument and service revenues. For 2014, with the higher mix of instrument revenue expected compared to last year, we are forecasting gross margin to be between 25% and 30%. As a reminder, at our current revenue levels, small changes in gross profit dollars can cause fluctuations in our quarterly gross margin percentage. Our operating expenses have been very stable over the past year, averaging approximately $21 million per quarter, and we plan on maintaining expenses at about this level throughout 2014. Please keep in mind that our quarterly expenses can vary due to the timing of certain research and development expenses. Our operating expenses continue to include noncash stock compensation expense and depreciation expense that, together, amounts to approximately $4 million per quarter. As Susan mentioned earlier, the net interest expense of $200,000 we recorded in the fourth quarter included a benefit of $300,000 from the revaluation of the financing derivative associated with the debt financing we undertook in Q1. As a reminder, the cash interest on the debt is $450,000 per quarter, and the scheduled noncash amortization expense is currently approximately $180,000 per quarter. Consequently, for 2014, we expect to record between $600,000 and $650,000 in net interest expense per quarter. With regard to cash, we continue to work on reducing our cash usage. In 2013, excluding the $75 million in cash we raised from a combination of debt and equity offerings and our collaboration agreement with Roche, we consumed roughly $63 million. At the start of the year, we mentioned that we were targeting to consume $70 million or less, so we are pleased with where we ended with the year. For 2014, we are targeting to consume $60 million or less in cash. And with that, we will open the call to your questions.

[Operator Instructions] And our first question comes from Daniel Brennan of Morgan Stanley.

So certainly, a strong finish of the year on bookings front and on the consumable pull-through. Mike and Ben and Susan, how are you guys thinking about the bookings outlook for '14, given the pipeline that you're seeing in front of you today?

Well, just a comment first on the bookings that we had in 2013. On a year-over-year basis, we think it's stronger because of a combination of things: that's the product enhancements, the customer publications, the extending applications. And we've always said that those kinds of things should drive higher bookings. So generally speaking, even though it's kind of soon [ph] to call it quarter-to-quarter, we expect bookings to grow.

And then I guess, Ben, maybe related to that, and then we can move on to just maybe another question on some of the utility of the RS today. So is it more of a function that you guys were confident to call it doubling or more, which turned out to be the correct call? Is it more of a function of just a level of conservatism? Or is maybe something with the pipeline or maybe the customer base you're going into don't give you as much visibility?

Well, I think one of the things we're trying to do is kind of focus more on the revenue side since that becomes -- as it becomes a bigger number, more important. Obviously, you have to get the orders in order for the revenue to grow. The orders rate on an individual quarter because we're still at relatively small numbers is -- it isn't so much as hard to predict over a longer period of time. It's really hard to accurately gauge it on a short-term basis. And so as we get a little more into the year, we know all the things that can happen out in the funding environments and different areas of the world and so forth. We'll perhaps give a clear vision of what the numbers are likely to be, but we feel pretty good looking at all of our sources of revenue with the guidance that Ben gave relative to the total sales.

Great. And then maybe on the Roche partnership. So I guess, you mentioned things are off to a good start, creating products. When can we expect more kind of color about how we should think about that partnership? And are there types of products that may come out? Or just trying to get more visibility or going to help thinking about kind of the impact of that partnership. I know you talked about initial kind of regulatory filings to date, 2 to 3 years. But should we expect more information, material information, as the year plays out this year?

Well, if you remember how we answered that in the last conference call, we want to be a little careful as does Roche in tipping our hands a little too early on exactly the kinds of test that they want to focus on, because that's competitive information that doesn't need to be out there too early. And as we get closer to understanding what the regulatory time line is likely to be, and that relates to a specific test that we would go after first with Roche, we'll clearly give you more guidance. But it's like some potential candidates for president in election 2.5 years from now, you don't have to announce until you have to announce. And we don't feel that we're under any particular push to announce too early.

And then maybe one more, in terms of -- to the extent you execute on your product roadmap again this year with a fourfold increase in the throughput with the read length and all the software enhancements. Does that increase -- does that open -- the RS open up more significantly to any of the markets at which you've discussed? Certainly bacterial genomes is kind of your bread-and-butter, but the ag market certainly is something you're growing in and human [ph], you've talked about now. So how should we think about the markets kind of that you're penetrating today versus what you could be penetrating as the year plays out?

Well, they certainly expand in terms of the number of people who are willing to look at the RS in some of those other areas. An interesting statistic relative to the presentations coming up at AGBT next week, there are more posters and talks on the PacBio RS being used for human sequencing than there are for bacterial sequencing, which is clearly a first for us. And they're almost as many in the plant side as there are in the bacterial side. So we're already seeing as the throughput has gone up and as the software tools have improved, as people realize how much they can pick up out of these larger genomes that they just weren't seeing with the short-read technologies, an increased interest in those areas. And so we would expect that as the results of some of these early studies become more widely known, not even with the current throughput, we're getting a lot of interest from additional people within those areas.

And then -- sorry, maybe one more. Just on the consumable pull-through, how should we think about that for 2014?

That metric is one that sometimes you don't want to just take a one-quarter metric. We mentioned that we're well over $100,000. And in fact, if you just took the fourth quarter, you can talk yourself into something like $125,000. Particularly careful with some of that stuff because if you just take it over a broader time period, it might average out a little bit. That said, generally speaking, we're working on increasing it. It's just sometimes you can't track it on a quarter-by-quarter basis. And that's why we gave you the metric more on a year-over-year basis.

Our next question comes from Bryan Brokmeier of Maxim Group.

Where is your sales force today? And can you talk about any plans that you have for changes to the size of it in 2014?

Brian, this is Ben. So salesforce is still roughly the same. We mentioned that we have about a total of 50 people in the field, and that's a combination of salespeople and support people and technical service people. That's actually been fairly stable for at least a year now. I'll have Mike just maybe color us up a little bit. And there are probably some spots where we're going to look to add some resources, but we're not planning dramatic changes at this point.

Yes, the thing that we've mentioned it before is that on the consumable side, we have been utilizing some of our field technical support people more in a consumables sales mode, not necessarily as quota sales reps, but to help particularly, groups like those that run core facilities, help market their services to the broader audience that submits samples to them. And that's been a nontrivial boost to the kind of reagent sales that we've seen in those laboratories. And so having done that on a bit of a pilot basis in the last year, we plan to expand that this year. It's not changing total headcount; it's changing some of the roles and responsibilities and some of the people we already have. And as Ben said, we will probably make selective additions in certain territories, on quota sales reps as well.

And is any of your backlog from your existing customer base there and how does that compare to the percentage of customers in your installed base that own more than one instrument?

Yes, I think I got the question. So none of the 9 new bookings were with existing sites. So those are going to be new sites. And we're not at liberty to say who the new purchasers are because those are usually under some sort of confidential things. But there was at least one multisale in those 9 bookings. Both are unit sales.

Okay. And you tend to see customers that own more than one instrument run at a higher utilization, right?

Yes.

And then lastly, in you operating expenses. You been trending at the low ends of where you've been guiding us for the last couple of quarters. What should cause that to move higher such as we see increased R&D as we near some of your Roche milestones? Or how should we think about the operating expenses?

Well, R&D actually did, let's say, jump from Q3 to Q4, a pretty fair amount if you kind of looked at it. So the development expenses, they can move a little bit because they tend to be project-based. And so without calling it exact quarter-to-quarter, I think it's fair that the way we're trying to manage it is somewhere around the same of $21 million a quarter. And of course, it's going to be a little bit higher, a little bit lower, depending on what quarter it is.

Our next question comes from Bill Quirk of Piper Jaffray.

It's actually Dave Clair here for Bill. So first one for me, I was just hoping, how should we think about the pacing of improvements to RS in 2014?

Well, I mean, we expect them to be continuous, but not necessarily at the same level each time we make an improvement. So we'd expect to do more than one software upgrade, which adds functionality to the platform, particularly, in terms of more push-button support for a lot of these applications. IsoSeq is one of the ones we've mentioned. We will roll out changes in our chemistry, and we will be continuously rolling out new protocols for sample prep improvement performance. So it'll be something that gets done multiple times in bits and pieces through the year. It's not one thing that suddenly jumps by itself up to 4x improvement. As it was last year or the year before, it was accumulation of several smaller steps that got us there.

Okay. And then I think you mentioned average read lengths, we should kind of expect like 20,000 bases in 2 years. Where do you think you can get in 2014?

Well, some reasonable chunk of that.

Our next question comes from Zarak Khurshid of Wedbush Securities.

In terms of the new deliveries, just curious how quickly customers are ramping up to kind of meaningful steady-state levels of consumable pull-through?

Yes. It depends a little on the customer. Some of them have projects ready to go. Others take a little bit longer in terms of training and getting the kind of projects that they want to put on it in -- that are at scale. So I would say, some ramp-up and we had one shipment early last year that wound up being one of our higher-usage customers almost from day 1. Others can take 4, 5, 6 months to really get up to a reasonable run rate. So it's kind of all over the place. Our goal is to try to get them up as quickly as possible, and one of the reasons that we have to sort of refocus over our field applications people is to help in that process. I said they're not necessarily quoted sales reps, but they're -- one of their goals is to make sure that the customers are ready to use the machine as efficiently as possible, as soon as possible. And that's helped us get the overall consumables pull-through per instrument going upward.

Understood. And then maybe a question for Ben. I think I missed it in the prepared remarks. But what were the expectations for the Roche revenue for this year?

So Q4 is a pretty good indicator. So it's scheduled to be amortized at $1.7 million per quarter, and so the guidance is that it'll be $6.8 million for the year.

[Operator Instructions] Our next question comes from Tejas Savant of JPMorgan. Tejas Savant - JP Morgan Chase & Co, Research Division: Just stepping in for Tycho here. Can you talk a little bit about the reasons NIH budget increased? And have you heard anything from your customers in terms of spending patterns getting a little bit better going forward? And how is that sort of factored into your 2014 guidance?

Well, it's certainly better than a sharp stick in the eye in the sense that it's not like it's a whopping increase in a longer-term set. It's about, what, the budget was 3, 4 years ago. That said, it's substantially better than the worst case that people are having to plan for. And more importantly, it's a little more secure in giving people the ability to kind of plan what their budgets are going to be over, say, the next 2 years. And that's the thing that's made sort of our customer base a little bit more encouraged than they were before. Getting NIH money is still pretty tough. The grant success rate is a lot lower than historical averages over the last 2 or 3 decades. But it's -- and I think customers are certainly more comfortable now in planning programs and the kind of purchases that can go along with those from equivalent perspective than they were 2 months ago. I'm positive, but early, I guess, is the question. Does that answer your question? Tejas Savant - JP Morgan Chase & Co, Research Division: Okay. And in terms of the strength in consumables in the quarter, can you talk about how broad-based it was?

Pretty broad, overall. I think we said in the past it's -- a lot of times, there's this rule of thumb where you think 20% of your customers represent 80% of your revenue. That's not the case here at all. It's actually much more broad than that. So we're pretty pleased with the sort of broad-based increases in utilization that we saw in the quarter.

Our next question comes from Daniel Brennan of Morgan Stanley.

I just wanted to do just a follow-up. Mike, I mean, since you joined the company, the company's had a pretty significant turnaround. And I think one approach that you've taken and it appears with Ben and Susan is to not set guidance on an annual basis or things that maybe are beyond your control or too much uncertainty and kind of get the street focused on, maybe, some of your long-term goals, where usually you have visibility. So just with that being said, just back to the guidance for '14, when we think about this 55% revenue growth that you've guided to, that looks to be like a really potentially low bar, given the significant growth in instruments that you have in the field, particularly with the consumable pull-through that you're generating today. So I'm just wondering, is the lack of bookings guidance or the 55% revenue guidance, which is certainly very healthy, but nonetheless, based upon the higher instrument number -- is that just predicated upon a philosophy of just trying to kind of set a low bar? Or is there anything that's potentially happening in the field? Maybe Illumina's expensive new product lineup is potentially something you're keeping -- an eye on given -- you just mentioned grant dollars Illumina -- just kind of put together the guidance for the top line with the bookings and kind of what you're seeing.

Right, I have to say one thing, Dan. I mean, if you look at Illumina's guidance, which is spectacular at 20%, we think 40% is even more spectacular. So I think we hope this is a bullish number. We are coming in the year with 13%, but we're coming off a year that we did 25%. So we're guiding up on all elements for the revenue, and it's meant to be positive in every way. Bookings are hard to call. We probably will see less bookings this quarter because of the seasonality. And when you're in the law of 1 or 2 numbers, it's wrong to get too far there. We've guided last year on the doubling at the end of the second quarter conference call. So I think we're just trying to give you numbers we truly believe in and that we stand behind us like we do with our -- what we deliver to our customers.

Yes, and I was -- I mean, I certainly wasn't taking away fromthe55%. That's a big number. It was just when we think about the pull-through on what you have in solitary -- that was all. I was just trying to kind of come back. Because I think that'll be an investor question. That's all, Susan.

So if you look at 88 installed base and even a large percentage, an increase in pull-through doesn't really drive against it, the amount of revenue we get per instrument. So that's a dominant issue still for us.

Yes, I think, reemphasizing the point that Susan just made, we did give you guidance last year on orders and bookings. But we didn't do it until we had really seen what was going on out there. I mean, a lot of things could happen. Weird announcements from potential competitors can have a short-term influence on what happens. Changes in budget priorities, not just in the U.S., but in Europe or Asia can give you short-term lack of clarity as to what's going on. And so we felt pretty comfortable with the sort of at least number on the revenue growth. And as we did last year, we're likely to give you a more updated number as we get a better visibility into the year. We are at still small enough numbers -- that trying to think we know exactly what's going to happen is kind of nuts. And we want to try to give you something that we think is a reasonable target for us to achieve. And our goal is to do as well as we can in the marketplace. And we feel pretty enthusiastic about the progress we've made and the opportunity we have. But clairvoyance is not something I've been able to develop over the last 30 years in this business, as well as I would like to. So we're trying to be reasonable, but to still convey a sense that we feel pretty optimistic about our opportunities.

And as you know, no more questions at this time.

Okay. That was our last question for the day. In closing, 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 a its superior performance and sequencing accuracy, uniformity of coverage, extremely long read lengths and the abilities to characterize DNA-based modifications. We continue to make progress into driving the adoption of our products and it's rewarding to see momentum building in our business. We look forward to seeing some of you next week at the AGBT, and we will talk again in 3 months' time.

Ladies and gentlemen, thank you for your attendance. This does conclude our conference. You may now disconnect.