Compugen Ltd (CGEN) 2007 Q4 法說會逐字稿

Welcome to the Compugen Ltd. 2007 fourth quarter and year-end 2007 financial results conference call.

(OPERATOR INSTRUCTIONS)

As a reminder, this call is being recorded, February 19, 2008. With us on line today are Mr. Martin Gerstel, Chairman and Alex Kotzer, President and CEO.

I would like to remind everyone the Safe Harbor language contained in today's press release also pertains to all content in this conference call. If you have not received a copy of today's release, and would like to do so, please contact Ronit Lerner at telephone number 9723 765 8560. Mr. Kotzer, would you like to begin?

Greetings and thank you all for joining our fourth quarter and full year 2007 conference call today. Ronit Lerner, our CFO is not participating in today's call and she has an excellent reason. This week, she's expecting to deliver twin babies. I am sure that you all join us in sending her our best wishes.

I will start today's call by highlighting a few financial results for the past year and then provide an update on our progress and achievements. I will then turn the call over to Martin Gerstel, our Chairman who will briefly review key portions of the new corporate presentation that has been posted on our website today. We suggest that you have access to the presentation in order to more easily follow Martin's remarks, which you can do by going to our website home page at www.cgen.com. After Martin's remarks, both of us will be available to answer any questions you may have.

I assume some of you familiar with Compugen noticed the change in the terminology used in our press release today. For some time now, we have realized the use of the term "discover engine" for our discovery capability in a specific field of interest could be confusing, since several engines could be combined together to predict and generate specific discoveries in therapeutic or diagnostic areas. In addition, the same engine can be used for different applications. For this reason, we have elected to change the terminology from discovery engines to discovery platforms, which we believe describes more accurately the capability.

With respect to Compugen's financial results, our reported results continue to be fully in line with the company's past guidance. Current revenues remain insignificant with revenues for the fourth quarter of 2007 of $90,000 compared to $10,000 for the fourth quarter of 2006, in both cases representing small initial milestone payments under an earlier agreement.

Research and development expenses of $9.7 million, compared to $10.9 million for 2006, remains the company's largest expense, representing more than 35% of total operating costs. (Inaudible) in 2007, we have a net reduction in cash and equivalents of $9.2 million, compared with our earlier guidance of $9 million to $11 million ended the year with $17.2 million. For 2008, the company anticipates operating cash requirements similar to the actual usage for 2007 with a potential small increase due primarily to the weakening of the U.S. dollar compared with the New Israeli Shekel.

Our ongoing development and business activities continue to expand to the new platforms and new collaborations. Since the beginning of 2007, we have announced seven new agreements. Targets for antibodies and drug candidates with Medarex, toxicity markers with Teva, a second immunoassay diagnostic market agreement with Biosite, markers for unstable atherosclerosis (inaudible) with the Mayo clinic, collaboration agreement with Teva covering server validation and licensing option of our MCP-1 therapeutical (inaudible), genetic markers predictive of drug response with (inaudible) patients with Roche and most recently, a discovery and license option agreement with Merck for GPCR ligands. Martin will explain our unique discovery platforms and how a small company like us can have so many projects in so many seemingly different fields.

But even without these understandings, the commercial potential embodied in these collaborations should be obvious. The most recent collaborations with two of the leading research based pharmaceutical companies in the world, Roche and Merck, are particularly important to us. This is not only because of the important market position of our [department] but also because they're based on two of our newest discovery platforms, both of which have significant opportunities for additional collaboration. Since the beginning of 2007, we have also announced the three therapeutic candidates who are advancing into our pipeline and two new discovery platforms were validated.

Since the focus of today's call is a review of our new corporate presentation, I will conclude my talk by stating we are looking forward to the challenges and expected achievements of 2008, which could be a year of substantial progress in many areas; further validation of our powerful predictive discovery capabilities, advancement of therapeutic and diagnostic candidates, potential for meaningful milestone payments and most importantly, a deeper recognition of the potential available in our company, both scientifically and commercially.

And now, I would like to turn the call to Martin.

Thanks, Alex.

Before going to the presentation, a few introductory remarks. As Alex mentioned, you will probably find it much easier to follow my comments if you have the presentation and as Alex mentioned, you can access it by going to the www.cgen.com home page and clicking on Corporate Presentation, February 2008 under Events.

Second, I'm not going to discuss the entire presentation and the parts that I will discuss will be reviewed very quickly. In particular, I will not be covering any of the slides providing supporting data for the individual product candidates that we have used in the presentation as examples of our capabilities in various areas.

We consider this presentation and today's call to be very significant in the evolving story of Compugen. In the past, I have made a number of very strong statements without being able to demonstrate much independent support for them, such as Compugen has an extremely powerful and very unique discovery capability, our uniqueness comes from our proprietary and predictive understanding of important aspects of the underlying science, Compugen's discovery capability is broadly applicable field after field, our capabilities get better every day. Because of our unique approach, compared to other companies that do not have the underlying scientific base that we have created over the past decade, in general, when we decide to focus on a new field, we're already 70% to 80% finished before we even begin the program. Although and lastly although it appears contradictory, our unique approach allows us to pursue many different fields in both drugs and diagnostics, while at the same time, remaining a very, very focused company.

In the past, I have also stated that I am confident that the combination of our broadly applicable discovery capability and our business model provides the opportunity to create a rapidly-growing company with extremely high profitability since our revenues will be largely milestones and royalties, and very low risk since we have so many products and so many different fields being developed at the cost and risk of many partners that themselves are leading companies in their respective fields.

As I mentioned, these are pretty strong statements; however, we are now seeing a growing number of validation events for both our approach to discovery and our business model. And, therefore, this presentation is aimed at providing our shareholders and others for the first time, a clearer understanding of how we do product candidate discovery, why we believe we have a significant competitive advantage and in general, why I have felt comfortable making the very strong statements I have made in the past.

After reviewing this presentation, you may or may not agree with me regarding the outlook for Compugen, but I believe you will have a very good understanding of what we are doing and how we do it.

And with that, let's begin to look at the presentation starting with slide two, which is the safe harbor statement which the moderator has already explained; it covers this, our call today. The slide numbers are marked very clearly on the lower right-hand side of each slide.

So, moving to slide three, I want to say a few words about traditional discovery, just to put things in context before we get to what we do. Essentially all the traditional discovery in the drug and diagnostic world has been based on observation and validation. The observation can be either accidental or more likely in a corporate environment, done by screening thousands, in some cases, millions of compounds. In the years past, it was done with rats and other animals. More recently, it's done with highly automated, high throughput robotics and very, very sophisticated technology, but it's basically, although you're getting many, many more observations that, you know, in a short period of time, it's still the same basic concept. And that is set up an experiment and observe and hope that you see something interesting and if you do, then you go ahead and validate experimentally.

If you move to slide four, there is only one thing that can you absolutely say for certain about this approach to discovery, and that is that it will, the productivity will clearly, has to decrease over time because by definition, you find the easiest things first and then it gets more and more difficult.

And you can make up for this increasing difficulty for awhile by increasing the automation, by doing more experiments, but in the end, it catches up with you and it has caught up with the pharmaceutical industry as is shown in slide five. This is the very famous slide shown at just about every conference that you will go to with respect to the drug industry showing how the research and development costs continue to go up exponentially, approaching $45 billion a year now and yet new drugs get less and less with less than 20 in the last few years.

Moving to slide eight, now, think, to put in perspective now what Compugen is doing, first you need to understand that we're, the only reason that we can do research and discovery the way we're doing it is because we have spent a decade building a very unique base. In the past, I mentioned for the first number of years in the company, we were focusing mainly on the science. I probably should have been more specific because it wasn't science in general. It was very directed. We were focusing our total efforts on how do genes express transcripts. How do transcripts become protein? How are proteins cleaved to create peptides, et cetera. And since these are the models that either are the therapy or the target for therapy or the biomarker, having this understanding, it was my view would clearly assist you in moving forward with respect to discovery. In parallel, as we were doing the science, we created an exceptional R&D team and we have what are well-recognized as the leading systems and schools in the world for computational biology.

Moving to slide 13, we, a few years ago, reached the point where we felt we had a significant capability that we could begin now to focus and actually start to go after making the predictive discoveries in various fields. The first field we chose was therapeutic proteins. We have a number of interesting discoveries there. As Alex mentioned, we entered into a collaboration with Teva for one of our therapeutic proteins. But we, shortly after we were working on the protein area, we decided to move into immunoassay diagnostics; and diagnostics by definition are somewhat easier to do than therapeutics and this program as shown on slide 13, moves very quickly so that by the end of 2005, we actually had not only made discoveries, but we had agreements with a number of companies.

What I would now like to do is to very clearly point out how we do our discovery. Every time we approach a new field, we do it essentially the same way. There are four steps that we take in moving forward, and I'm going to give a couple of examples in different fields. The first one is perhaps the easiest one to talk about and that is the field of immunoassay diagnostics and this is on slide 14.

Moving to slide 15, whenever we start a new field, the first thing that we do is that we utilize our understandings of how genes express transcripts - how transcripts become proteins, how proteins become peptides and our predictive modeling of these very, very complex biological phenomena, in order to predict an in silico inventory of possible product candidates. We use our basic understanding of how the science works at the molecular level to create in our computers, hundreds of thousands, many cases millions of possible answers for whatever question you're, whatever field that you're looking at.

Now, as I said, the example that we're using of immunoassay diagnostics is very easy to say what this in silico inventory of possible product candidate should be, since every, on slide 16, since it shows that since every marker that is used in immunoassay diagnostics is a protein. So, therefore, what you want to have as your in silico inventory is a listing or a collection in the computer of every possible protein that could exist in the human body.

Now, in view of our understandings of -- on slide 17, in view of our understandings of how genes express transcripts and how transcripts become proteins, we have modeled this. So, in fact, we do have now a in silico model of the human proteome.

You move to slide 18, as I said, all of this work is done in the computer. As of now, there has been no experiment, lab experiments done with respect to this, with respect to this field. The second step that we take when we approach a new field is now to select candidates from this in silico inventory that we believe have a high likelihood of being a successful product candidate for whatever field you're looking at. And here, the main, our main advantage is our computational tools, our ability, our algorithms that we have created already and our ability to create other algorithms and an understanding of how this all works at the molecular level to be able to create the set of tools that will then look at this entire inventory of possible answers and pull out those that look reasonable as shown on slide 19.

Slide 20, now, with respect to the example that I am using, immunoassay diagnostics, on slide 20, I have listed there some of the types of things that you would base your algorithms on in order to find these markers that could be useful for immunoassay diagnostics. For example, differentially expressed spliced variants, tissues with specific expression or antigens that are already known to be disease-associated. We, we have done that with respect to this field and the result was, as shown on slide 20, hundreds of novel spliced variants of known diagnostic markers.

The third step in every discovery platform is shown on slide 21. The -- here what we do is now we move to the laboratory, to the wet lab and in most cases, we use standard conventional in vitro and in vivo tests to validate whether our in silico predictions are correct or not.

If you look to slide 22, in this situation, immunoassay diagnostics, we use the typical Elijah and the realtime PCR programs to, or capabilities in order to check out, in order to validate or not validate these in silico discoveries and as it turns out, we end up not with a discovery but with multiple candidates with many candidates that actually have now were initially predicted in your computer and now have been validated as not only existing but doing what you had predicted them to do.

In this case, we have -- on slide 23 -- you can see that we have entered into licensing agreements with a number of the major companies in the field for immunoassay diagnostics.

Moving to slide 24, you see one of the, perhaps, in some ways, the most powerful aspects of our approach to discovery, and that is you don't just get the multiple candidates when you have finished doing this activity. You now have new information. You now know that certain of your predictions were correct, certain predictions were not correct, and also during that period of time, have been more science, more understanding about this field developed either within our company or outside of our company. So you have the opportunity to take all of this information now and add it to your prior knowledge with respect to this field and run the engine again and that is why we say it's constantly improving. You run it once, you get the results, you can run it again and get new results. And actually, we have done this.

If you look at slide 25, this is the fourth step in the business of having these discovery platforms, the fourth and final step is that once it's created, you continue to do these repeat discovery runs that continue to provide you with new product candidates.

Coming back to our example of immunoassay diagnostics, we actually did that on slide 26. We -- one of our partners, Biosite, had a specific interest in a certain type of marker, so we modified the algorithms to go specifically after that type of marker, we ran the engine again, we came up with a whole group of new product candidates and they have also been licensed now to Biosite.

Slide 27 summarizes this in one slide, and if there is one sort of slide that, if you want to understand what we do as a company, this is the slide that you should look at and continue to try and understand. Because once you understand slide 27, you will have a sort of very, very good understanding of what we do, how we do it and why we believe it's both unique and very powerful. What you see on this slide is the issue that the discovery platform itself is made up of these three steps of prediction in silico, selection in silico and then a third step of validating in the laboratory in vitro and in vivo. And what allows this to happen is this core, the people, the science and the systems and tools that we have developed, that are available to us when we decide to go after a new field. And that is why I said earlier that when we start a new field, we're 70% to 80% done with the program before we even start it because of this core of understanding of the science that we bring to it. And then every time we run this platform, we end up not only with multiple candidates but we ended up by new definition, with new information that allows us then just to do it again.

Now, as you look at slide 29, by the end of 2006, basically during 2006, we started another eight, the development of another eight very interesting discovery platforms. You see the nucleic acid diagnostic, monoclonal antibody drugs where we have now the joint venture or joint collaboration with Medarex, the preclinical toxicity markers; we're working with Teva, drug response markers, this is the basis of the Roche agreement, GPCR-based drugs, this is the basis of the Merck agreement; drugs of new indications, which is our most recently announced platform and you will be hearing more about that in the coming months. And then there are a couple of other programs in the area of peptides that we haven't fully disclosed yet. You have can see this growing inventory of these discovery platforms that once validated, they continue to not only exist, but continue to improve with time.

What I now would like to do is to move to a second example, a much more complex example, much more complex than the immunoassay, from the standpoint of really showing you the power of the predictive capability that has been established at Compugen. This is on slide 30. What we'll be looking at are the GPCR-based drugs.

GPCRs are extremely important in the world of medicine, probably close to half of the drugs used in the world today work by modulating this one family of receptors. The history has said that if you find a new ligand for a GPCR, there is close to a 50% chance that it ultimately will become a drug, just about every major pharmaceutical company has a team of people trying to find new ligands for GPCR.

So, looking at slide 31, the question is -- so how do we approach this? This is a new field now, we're going to go after it. What do we do? Well, the same way with the immunoassay, the first thing that we do is we need to predict an in silico inventory of possible product candidates. Well, GPCR ligands in the body are typically peptides and so it's kind of, it's straightforward what you would like.

If you move to page 32, you would like an in silico inventory of all of the peptides that could be possible in the human body. That is easy to say, but the question is how do you find or create such a thing? Peptides sometimes can exist only for a second. They're very small. It's essentially impossible to do this through an experimental basis. So, what have we done? Move to slide 33.

As I showed you when we were, when I was talking about the immunoassay diagnostic program, we, by having a predicted human transcriptome, we were able to create a predicted human proteome, which was the basis for our immunoassay diagnostic program. Now, we have taken that predicted human proteome and we have utilized our knowledge of what distinguishes a protein, what makes a protein highly likely to actually be secreted into the blood stream. And so we have created utilizing that knowledge, we have created a predicted secretome all of the proteins that exist that should end up in the bloodstream, which, therefore, make them potential candidates for an immunoassay diagnostic.

Peptides, what peptides are, pieces of proteins and the way peptides are created in the body is that they are cleaved. They're cut by enzymes. The protein is cut by enzymes into smaller pieces, which are the peptides. So, what we did was we created a machine-learning capability that looked at all of the known protein cleavage sites, places on proteins where it was known that they were cut as it turned into peptides and we use this machine-learning capability to generate unknown novel, a prediction of unknown protein cleavage sites. Once we had that, we then combined that with our predicted secretome and ended up with a predicted human peptidome. Now we have what we need for our in silico inventory -- an in silico peptidome. If you look at this, it really is quite astonishing. Each one of these steps is an incredibly complicated complex, biological activity with a gene expressing a transcript, the transcript becoming a protein and what you have here is all in silico.

You look at slide 34, all of this is in silico and you have prediction on top of prediction on top of prediction on top of prediction. As of now, there is no way to know whether what we call an in silico peptidome, whether it has any validity or not. At this point in time as far as anyone knows, it could and probably most people would expect it to be total garbage at this point.

If you move to slide 35, the next step is, as with every engine, I'm sorry, every platform; I have to get used to the new terminology that I was just talking about. As you move to the next step of selecting candidates from this in silico inventory, the question is, okay, now you have hundreds of thousands of peptides. How do you pick out from that massive number of peptides, how do you pick out those that are likely to be ligands through GPCRs?

Well, if you move to slide 36, here we've created another machine-learning capability that looked at peptides that were known to be GPCR ligands and another group of peptides that were known not to be GPCR ligands. And this machine-learning system then identified what are the attributes that you would see on a peptide that you had seen for the first time that would lead you to believe that that peptide either was or was not a good candidate to be a ligand for a GPCR. We then took that information and applied it against our in silico peptidome and ended up now with novel peptides that are predicted to modulate GPCRs. So now, on top of all the other predictions, prediction on top of prediction on top of prediction on top of prediction, we now have three more prediction on top of prediction on top of prediction and we have this set of peptides. Again, that's slide 37, all in silico.

Not a single experiment has been done at this point in time. But now we move to the validation, slide 38 and we now go to the traditional in vitro, in vivo validation of these things. And the way, if you move to slide 39, we, at this point in time, we had somewhat over 100 potential candidates, over 100 peptides. Novel peptides, peptides that were not known to exist that we predicted should exist and we also predicted should be or had a high probability of being a ligand for a GPCR. We took a sample of 33 of these and experimentally went through validation and we found out that almost 25% of them actually were ligands for GPCRs, which is, well, it's quite an astonishing result and which probably means that within that other hundred, more than a hundred that are still left, there are probably others that we will identify over time. And it's from this eight that we have made our recent announcements with respect to further validation studies where we actually not only showed in vitro that they modulated the GPCR, but we also showed in vivo that you got the type of reaction either for cardiovascular disease or whatever you that were looking for that you actually, it did actually do what you wanted it to do in the in vivo, in the in vivo model.

If you move to slide 42, here again, now we've finished this program, at least the first run. We've created the platform, we've validated it and this actually. These peptides were actually just part of a validation run, a pilot run, to see whether or not the platform worked properly. And we ended up getting these, having these discoveries.

We are now going to be repeating this activity as I mentioned, as we had this opportunity to do with the new information, but what we'll be doing now is we'll be doing this, if you look at slide 43, we'll be doing the next run of this will be done to identify ligands for candidates of interest for Merck and this is the recent announcement that we made about our collaboration with Merck.

Move to slide 44, I am going to do one more example of, because I find this one to be really quite astonishing from the standpoint of the flexibility that this approach, that this approach gives you and that is our latest engine, which is the drugs with new indications.

So if you move to slide 45, the -- first let me comment that this is a very hot field now. Just about every big pharmaceutical, even smaller pharmaceutical companies, have one or more drugs that they would very much like to find a new indication for. The reasons are obvious that you know it's safe but, you know, it's out there in the marketplace and if you can find a new indication, then, of course, you can get additional patent coverage and all kinds of good things. And given the great difficulty in finding new drugs through the traditional route as I showed you in the beginning, this is a very, very hot area at the present time.

Now, what do people do? What does everyone else do if they're trying to find a new indication for an old drug? As I said at the very beginning, traditional research says you observe and then you validate. So, what everyone is doing is that one way or another, they are taking their drug of interest and putting it into multiple disease situations, either with cells or with animals or with all kinds of different formats, but the bottom line is to create, to try your compound out in many, many different settings to see whether it might light up in another disease, the disease situation. That is what everyone does.

So what do we do? Well, as with all of our fields that we enter, we start by identifying the in silico inventory. That is to say we want to have in our computers essentially an inventory of possible answers. So, what would be the in silico inventory for drugs with new indications? On slide 46, well, what we decided the in silico inventory would be essentially would be all drugs in use or development and all diseases. Now, obviously, there is no such thing as all diseases. Nobody, you know, every day they're finding new ones or breaking one disease down into others. So, the way we have handled that is our scientists have identified 1400 different conditions that could exist in a human body, it could have to do with a disease or a tissue or whatever, but a condition that you can identify and that if you, if this condition exists, then there is a high likelihood that it's associated with a certain disease. You may have condition eight, which is identified with asthma, condition 47, which is a certain type of cardiovascular disease, et cetera.

What our people did was they took essentially all the information that they could get available to them from the public areas. There is an enormous amount of information out there about drug targets, protein networks, netline abstracts, chick experiments; took all of this information and compared everything with everything -- making no exceptions. No judgements were applied at all, it was -- just take all the information available and compare everything that is out there with everything else. And the net result of this is that we now have in our computers for every drug target, every network, essentially every drug, we have them ranked from one to 1400. Each drug is ranked from one to 1400 with respect to these different conditions. You may have drugs, you know, drug one or drug A, that has, you know its first condition that's something that's known to be a CNS condition, the second one is also known to be a CNS condition and the third might be something entirely different.

Moving to slide 47, this is a phenomenal example of the type of capability taking all of this data from different sources and different formats, normalizing all of it and being able to compare everything to everything and the unbelievable amount of data here we call this our med platform. It's all done in silico, and as I said, the results of it is for every drug, we have it ranked from one to 1400. And now we're, if you move to slide 48, we're now in the process of going through this and selecting potential product candidates and also doing IP evaluation. That is always a part of the second step -- from out of this large inventory of possible answers, in silico, again, using only computers, pull out those that look good and make sure that they are okay from an IP standpoint.

In this area, look at slide 49. You have to be concerned about, and obviously, you have to eliminate all known indications. The way we do it is we not only find new indications, we obviously identify all the known indications. We have to eliminate those, we have to deal with the patent, the marketing status. And also, because people make mistakes and, you know, there are errors out there in the literature or in the experiments, that we look for multiple support for a new indication. The fact that there might be just one hint that something is a new indication, for now, we're ignoring it. We're looking for things where there are multiple supports.

Moving to slide 50, what we, at this point in time, we have identified hundreds of drugs with predicted new indications, which we are now taking in on a priority basis through in vitro, in vivo validation.

Slide 51, to date, we have taken four or we have had four predicted new indications validated in vitro and we have got the in vivo studies now ongoing.

Move to slide 54. So those are three examples in very, very different areas, drugs, therapy, diagnostics and very, very different issues, but all falling into the exact same format of utilizing our basic understanding of life at the molecular level to create this in silico inventory of possible answers. Then utilizing our computational biology skills, capabilities to select out from that pool those that have high probability and then with that and all of that in the computers and then you go to your lab and you check out the ones that look best and as can you see, we're having an extremely high success rate in what we're doing.

Moving to the business side of the company, what we intend to do and what we are doing is we have set our objectives as to maximize the number of products in the development pipelines of drug and diagnostic companies worldwide under milestone and revenue sharing licensing agreements with us. We want to be and believe that we can become the number one company in the world licensing into the pharmaceutical and diagnostic world and that, you know, in a short period of time, we expect a number of major companies to be essentially dependent on us for their pipelines.

Moving through slide 55, we have set our own internal targets that in three years, we would want to have at least 20 products in development in active development plus, an additional five, minimum of five drugs actually in clinical studies moving forward.

If you move to slide 56, it's sort of -- it's just looking at averages. You say well, what could this mean for Compugen from a dollar standpoint? Well, the milestones are, given our current burn rate, the milestones are meaningful. In the long-run, I don't think the milestones would be very meaningful to us because, obviously, the royalties will vastly overrun them, assuming that we are as successful as I believe that we should be. But, if you look at comparing diagnostics to therapeutics, that, and, again, we're not in a position and it would not be in our best interest to ever describe specific financial terms in any agreement. We're going to work broadly with the industry and the last thing in the world we want is to be talking and publicizing what specific financial terms are for one, you know, with one client company.

But, we can talk in generalities and in general, our diagnostics agreements should provide for milestone payments over a number of years as a cumulative number in the low millions of dollars. Therapeutics would be somewhat higher and could be quite substantial, but the very substantial numbers would come later on in the process -- Phase III or NDA, et cetera. The more important aspect of it, at least long-term is the percentage share of the revenues we would get. And we're entering into two types of arrangements; one, joint products where we actually share the developments and share the revenues. As of now, we only have one arrangement there and that is in therapeutics with Medarex, which is a 50/50 arrangement. Everything else that we're doing as of now, and I assume most of the stuff that we will be doing moving forward will be on a licensed product basis where we end up getting royalties. And in general, the royalties, what we're seeing, the royalties will be higher in diagnostics, you know, in the range of ten plus or minus and in the therapeutics, in the range of five plus or minus.

However, in general, again, we're talking averages here, product revenues on average therapeutic product revenues are much greater than on an average diagnostic. And just sort of looking in generalities, talking in generalities, you know, a decent diagnostic, you think in terms of 50 to $1 million to $250 million a year of therapeutic from the hundreds of millions into the billions of dollars per year. So when you then look at that with respect to our share, what that says is for every successful product that is out there, in addition to the milestones, we should be getting $5 million to $25 million plus per year from the diagnostic activity programs or products and on the therapeutic side, $10 million to $50 million or greater, again, not, you know, now you're really talking blue sky with respect to what will happen ultimately with these products.

Moving to slide 56, 58, just to put that in perspective, if we now look at our current status, where are these programs and what we believe the three-year potential for these current products should be, can you see it's far in excess of the 20 products in development and five in the clinics. So, we believe clearly that the potential is there to more than meet our objectives, but we're still -- we're at an early stage, so time will tell.

Moving to slide 60, just very quickly our financial status. As can you see, we burn about $10 million, the last in '07 and next year, it's closer than $9 million. We have about $17 million in cash in the bank, slightly less than two years of cash as of the end of last year. Our operating expenses are somewhat greater than that because of various types of programs that we are part of and we, of course, have no debt.

Moving to slide 61, just really as a summary, there is some other information about the company there. I think that, you know, we're a small company, 75 employees but almost all research, more than 2/3 of them are in the research and highly experienced, multidisciplinary and the, with our discovery platforms now, we believe that we have a capability that really has an advantage of orders of magnitude efficiency over the traditional way of doing things, clearly broadly applicable and by definition, every time we do it, they just get better since they're all hypothesis-driven. We also have an affiliate, Evogene; we -- we have currently, though, only approximately a 12% equity interest in that since we have not invested any further funds there; it's been, funded by third parties. But Evogene is using our predicted biologic capabilities in the ag-bio field and you'll be hearing a lot more about them because in the ag-bio world, these predictive capabilities can get to the market much quicker, at least can be fully validated much quicker than they can be on the human side.

I think that's it. That is the end of this short presentation. I hope somebody is still there. Let me just open it up now for questions, not only, obviously on the presentation, but on the results from last year or with respect to anything that Alex was talking about. So it's open for questions.

Thank you.

(OPERATOR INSTRUCTIONS)

The first question is from Jeffrey Grossman of Berenberg. Please go ahead.

Hi, Martin and Alex, good afternoon, I suppose there.

Martin, during the third quarter conference call, you indicated the company is a party to a relatively large number of negotiations with potential collaborators. Is this still the case?

Yes and that will continue to be the case, but I don't -- because we're in negotiations with a lot of people doesn't necessarily mean that we're going to sign a lot of deals. I believe we are, but from my experience, long-term experience in this field, what I realized is that if you want a good arrangement, sometimes they take a lot longer than you want, than you would like them to.

But yes, the answer is yes. We're in negotiations, discussions with many companies.

So, I understand from that there is a high level of interest on the part of big pharma in the companies; for example, therapeutic candidates, the recently announced GPCR peptide candidates.

I would say there is clearly a growing interest. I don't know whether I would really at this point call it very large, but clearly the trend is straight up in the air. I mean from not having people answer the phone not that long ago, now they calling us. And so it clearly, you know, the word is getting out there that there is something different going on here, but we, by no means, is most of the industry sort of after us now to do deals. But we have enough on our platter that we're quite confident.

Tell me then, what is the likelihood of us seeing Phase I or II during 2008?

Highly unlikely unless it was in the new indication period. New indications is a wild card because can you go from, you know, very, very quickly from the predictive stage into Phase II and Phase III. If, assuming that the new usage is going to be within the dosage range of the current usage, you know, that you can move extremely quickly. If you are leaving aside the new indications area, I don't think it's reasonable to expect anything, Alex, would you agree? It's hard, hard to imagine. I think in 2009, next year, we should start to see some stuff entering the clinic but not this year.

It was reported that Merck-Serono will cooperate with as many startup and developing pharmaceuticals and will later be able to buy the rights to buy the drugs or help market them around the world and that the Israeli Industry and Trade Ministry will finance half of the research and development costs of these companies. Is Compugen seeking to take advantage of this opportunity?

I would answer. First of all, as an ex-Serono employee, naturally I know that people there. I was participating in the celebration with the chief scientist and we forward them some interesting ideas and once we will have an agreement, we will let you know, of course.

Thank you. Last question concerning the GPCR peptide discovery engine. The first one resulted in 33 peptides, of course. Has there been a second one resulting in additional GPCR peptide ligand candidates?

First, you have the same problem I have. It's not an engine anymore, it's now a platform.

Platform, excuse me.

I know, I have the same problem here in using that term so much. And at one point, I wanted to call Compugen the engine company. I am glad nobody bought into that argument. We would be in big trouble now. As I mentioned, the next run of this is going to really be with the Merck activity, as far as I know. Alex?

First of all, the first round will end up with 33 peptides. It ended up as Martin mentioned, out of them all, we selected arbitrarily 33 for the validation. Like we do in most of our platforms, we can select only a few to validate themselves in the platform. The next round, we probably will generate in silico again, hundreds and more candidates and then we will see how to proceed.

But there has been an improvement, you're saying in the platform?

Yes, Martin even mentioned one of them. The second round will be more focusing on specific GPCRs where the first round was a general approach. Just looking for peptides that theoretically could be ligands for GPCR in general, our efforts today are focusing on more specific discovery.

Thank you very much.

Thank you.

(OPERATOR INSTRUCTIONS)

There are no further questions at this time. Before I ask Mr. Gerstel to go ahead with his closing statement, I would like to remind participants a replay of this call is scheduled to begin in two hours for a period of 72 hours. In the U.S., call 1-888-326-9320; in Israel, 03-9255-900. Internationally, please call 9723-9255-900 Mr. Gerstel, would you like to make a concluding statement?

Thank you.

I'm sorry, we actually have two late questioners. Would you like to receive them?

Yes.

The first one is Bill Chapman of Morgan Stanley, please go ahead.

Good afternoon, everyone. Can I ask about the toxicity marker arrangement with Teva? How is that going to date?

It's a little complicated to answer, but I can answer that it's going according to the programs and we're happy with what we're doing. Is that a reasonable answer?

It's very difficult for us to give, to really respond on specific programs where we're working with another, with a partner. But we have -- let me say we're very happy the way our programs in general are going forward. I am sorry we can't be more specific.

Okay. If, just generally speaking, this is something we have a shorter time horizon -- is it possible if all goes well that you could be marketing this marker to other firms this year? If all goes well, theoretically?

You know, Alex?

I'm not sure we can say this year, but we will try to target it. It depends on the validation of the findings that we have done.

That is not an-- we can't project it, but it's not impossible.

Does this have the potential, I mean this could be licensed to dozens of companies, I presume, if it's successful. Is that correct?

It's true because here, it could be used by many pharmaceutical companies because. It is not competing among them, so each one of them could use it for its own internal needs. So, you're right.

And do you see yourself licensing it or will you do collaborations with a testing service?

We are considering at the moment, both ways.

Okay.

We might be able to provide the service for doing this, the test itself via a level contractual or licensing to those companies that will be interested.

Okay and I'm curious, Martin, on your platform for the peptides that you have not disclosed, why is that?

I am sorry, you mean the eight.

On your new platforms?

Because they're not validated yet.

Okay.

We don't disclose platforms until they validated.

Okay.

Hopefully you'll be hearing more about one of them at least this year, I hope. But they're still in development.

Okay. One last question on investor relations. I find the company's science to be A-plus but lacking in investor relations. Will you be having -- are you going to consider someone here in the U.S., an investor relations firm to make it more convenient to discuss the company's matters or -- what is on the rise in investor relations?

We're having questions internally. First, let me say we're going to do more in investor relations. Whether or not that will involve a firm in the United States or some other mechanism, we're thinking about various things. We purposely have done nothing, only because, you know, it was so difficult, you know, telling the story without a lot of external validation. Now that that external validation is coming in, we feel that now is the time to really let people know what is going on here and that, obviously, was the reason for why I went through what I did with respect to the presentation and why we put it up on our website. I mean hopefully it answers a lot of questions that our shareholders have been asking us about. In the past, we have been using general statements about, you know, are we science-based or use engines, whatever. Hopefully now, you know, people will understand exactly what we were doing and they can make their own judgement as to what degree this is going to be successful.

Okay. Thank you very much.

Thank you.

The next question is from Brett Rice of Janney Montgomery Scott. Please go ahead.

Good afternoon. With the tremendous potential that Compugen has and that you have spent quite a bit of time running through with the slide presentation, why haven't we seen one of your big pharmaceutical partners perhaps stepping up and taking a big minority stake in your company to kind of tie up this technology that has such great potential?

Well, first we're not seeking an ownership change or whatever. The other thing is that we're -- we want to be a service provider to the industry. I mean I went through this exact same thing with my prior company, ALZA, where it was, we were essentially so much more valuable as an independent company than being a captive of one company. We have an amazing ability that is just going to get better and better for early-stage discovery and we want to feed the entire industry. Given that that is our objective, we're really not, we don't really represent a risk to any pharmaceutical company. We represent an opportunity. We represent something where, you know, there is no need to own us. So it's -- really, you know, I can't say it's not going to happen, but I hope it doesn't.

Great. Okay. Thank you and continued good luck.

Thank you. What we really need is for there to be more understanding of what we're doing so that people can be in a position to sort of, as I said, make up their own minds. I think if people have all the information, we're going to do very well, both in the commercial world with our, the pharma world and with the financial world. I think the issue, though, is getting exposure, having people really understand what we're doing.

Right. Thank you for taking my question.

We have a followup question from Bill Chapman of Morgan Stanley. Please go ahead.

Yes, thank you. On the $90,000-milestone payment you got in this last quarter, that is for a second diagnostic product that has gone to the next level?

Let me first say that it's not really a $90,000 payment. Under accounting rules you have to take stuff in over time, so don't assume that if we report, and as we said last year it was $10,000 in the quarter, this year $90,000, don't assume that is one $10,000 fee and one $90,000. It doesn't work that way under accounting rules. You have to spread it out over the period of different, based on different rules, whatever. Let me just say that all of the amounts shown, payments that we received to date are all related to steps that are very, very early in the process and are really minor, relatively minor and are all involved with diagnostics. Is that true, Alex?

Could you tell us at least how many payments apply -- this applies towards?

How many payments from different -- .

Yeah, like three different, two different tests, molecules?

It's from one agreement. I would not mention for how many candidates and as you said, they are agreed milestone. This doesn't mean this is the beginning; could be the finish of accepting the agreement. They are all from the same agreement.

First, I am glad that we're getting some milestones only from the standpoint that it shows us something is happening. On the other hand, they're so minor, I would really hope that people don't sort of, you know, build anything into them. Hopefully as Alex mentioned, hopefully this year we're going to start to see meaningful milestones. Time will tell. Milestones are milestones.

Okay, one last question. Is there a possibility that someone could pay you to do certain platform-specific testing that could help us cut down our burn rate?

Yes, but I'm not sure I would want to do it. At one point, when we were building the company, we did some of that where we took certain of our capabilities and analyzed other people's data for them and whatever and made a few million, you know, three or $4 million on a program with one company they paid us $10 million over a three-year period for us to analyze their data, basically for them. So, there is that opportunity but we're not getting the value of what we're providing because we really have really enabling technology here and I really want us to continue to sell or license the results of using our technology rather than sharing our technology with others.

We are trying to get our partners paying for this specific effort we're doing for them but, of course, this comes on top of the milestone and the royalties which should be the basis of the agreement but as Martin mentioned, when we are making a run for partners, 70% or 80% of the work on the platform had been done, so relating to this cost to have a run, a specific run dedicated for the needs of the customers are relatively low. As I mentioned, one of our objective is trying to get them from this type of activity on top of the milestone and royalties.

Okay. A couple of conference calls ago you motioned seven diagnostic arrangements, four have shown validity. Where are you at on this process at this point? In terms of success?

With respect to the diagnostics, they have to go through a number of stages. Keep in mind that diagnostics we license out much earlier than therapeutics because the companies are willing to take them much earlier because the pathway it's a lot faster. But the first thing you have to prove is that they exist, actually we do that. We do the prediction and we show that they really exist, but then you have to show that they're measurable in blood, that they have differential expression, that they're whatever. I mean a whole bunch of steps and so we have molecules that are at their stages moving along in those steps. The nice thing about diagnostics is that once something gets over kind of the key hurdle, it can get to the marketplace very, very quickly.

And I will say now so I don't raise any improper expectations, but I doubt if we'll be able, we'll be willing to keep you up to date on this. As of now, none of our molecules have sort of gotten over that hurdle point of saying aha, it's on its way to a product, all right; they're still at earlier stages. The nice thing, as I said, is once they do get over, if they do, then the path to the marketplace is fairly quick and fairly secure.

Okay. Thank you very much.

The next question is from Jeff Gilbert of Peak Investment Partners. Please go ahead.

Good afternoon, gentlemen, congratulations on the year and thank you very much for the presentation.

Thank you for saying that. I greatly appreciate that. I really do. Thank you.

It helped me a lot. I'm not with a background in the field, so one question, if you could comment on the protection of your intellectual property, both on hardware perspective and also on personnel.

Let me just say that is a great question because, I mean that is what we have here.

I believe an intellectual powerhouse with respect to the specific areas that we are working on. I don't think there is, I think we're out here in the world with respect to this specific areas that we're dealing with.

The main protection that we have from a traditional intellectual property standpoint covers the discoveries that we make themselves. We don't show anything to a partner before we file some form of protection and whatever. With respect to just a general capability within the company, very difficult to protect it in any way. It's, you know, but it's not a question of being in the head of one person or two person. We have gone far beyond that now and it's so complex and so multidimensional now that it's a core of people. It's 25 of our key research people left, I would worry about the future of the company. One or two we can deal with.

It's, but it's not an issue that we have a secret formula for Coca-Cola here. It's just a massive amount of understanding of how life works at the molecular level; and everything we do there is teamwork. Teams of people, biologists and people who were formerly physicists and mathematicians who are now essentially theoretical biologists, people coming with different areas of the life sciences and what comes out is sort of the result of bringing together all of this, these different perspectives and capabilities and different knowledge. And so I guess the answer is one, we don't have a lot of protection from our IP standpoint and what is valuable here, our core capability. But as long as, I'm -- you know, we do what we can do and I think -- .

I understand from a personnel perspective. How about the hardware. Obviously.

That is a lot less important.

Okay. Sounds good. Thank you, gentlemen.

Thank you.

Thank you. There are no further questions at this time. Mr. Gerstel, would you like to make a concluding statement?

Yes, thank you, I appreciate everybody listening through my monologue. I hope it was helpful.

We are an unusual company and so we had some debate here as to whether or not it was a good idea or bad idea to go through that, but I think we kind of owe it to people to at least give them the information. And with that in mind, I would urge people if you found that at all interesting or whatever, I would urge you to ask actually now go through the whole presentation. I think with the discussion that we had hopefully, you know, it will now make sense and you will be -- there is a lot of additional information in there about validation issues and more details and backups. I would encourage you to do that. And also, if we have considered actually having another conference call at some point in the not-too-distant future where we actually would go through the entire presentation and open it up to having more of a discussion of the presentation. Not necessarily up, of the issues raised in the public, in the presentation. And we have not really made up our minds as to whether or not we're going to do that or not. If people out there would be interested in participating in such a thing, just drop us an e-mail and if we do it, we'll let you know. So, other than, that let me just thank everybody for participating today and I see a number of our really loyal shareholders were on the line and we really appreciate your long-term belief in us and we think we're now about to deliver and we appreciate it. Thank you very much and good night.

It's a good night.

Thank you, this concludes the Compugen Ltd. fourth quarter 2007 year-end results conference call. Thank you for your participation. You may go ahead and disconnect.