WOS 2 / Proceedings / Panels / "Geistiges Eigentum" / Informationsvielfalt: Eigentum an Informationen / Diskussion / skript


Question: I had a comment on Tim's final remarks on annotation. And one reason perhaps that a lot of providers of maps ect. may not be very keen on annotation is the value proposition. Even if these are public sides or sides which are not commercial and their are providing their information in public. If I provide a service which lists the best bars in Berlin and I take a reference map of Berlin the value by the enduser the value that is being the highest value that is being perceived that I have added. But in terms of cost and effort the value that is going in to the final thing that the enduser sees is the people of who are serving up the map. And so I get the credit or I get the money or I get the advertising whatever. But regardless of how the value is paid for I get the value for adding the last bit, the annotation and it is difficult to model how the people who come up with the highest cost get value back. And that might be a practical problem I think. It is not as if you don't have standards. Each (term) in any case I perfectly made for annotation even without having a different browser. On your webside you have links to pages elsewhere. You have links to frames on other sides and other servers. It is perfectly possibly for you to do that with ordinary html technology in ordinary webside. The reason that it doesn't happen is that people don't provide their data in that way because they want to have control with the enduser and excess with the enduser.

Tim Hubbard: That doesn't invalidate this existing html models at all. It is simple an alternative. Infect these (dails) systems we have actually integrated it on our webside. So, rather then download on ever browser you can also see our annotation and add other server to it directly, rather then to deal with another interface. But if you got a specialist (dials) viewer then you can link back to the original side that provided the data. These things about you know annotation giving value, value needed to be paid for, paying is one of the mechanism for advertising, all these things can be supported. There is no reason why you cant have dial servers that are private that you subscribe to. You only see that bit of annotation if you pay for it. But anyway this is open source and the basic of open source is that people do things and give it away for free. So, you can try setting the revenue stream and you may be competed against by people doing it for free. It don't think it even invalid any of existing things what most of your complains can be put in to, can be justed into the model.

Perhaps this is also a question for (...) as well in terms of what the bioinformatics industry looks like these days because it seems like on of the proposition made in that industry is precisely providing that kind of (frontant) or integration of data for pharmaceutical companies.

Let me give you my worst case scenario for companies. This is the aggressive public domain policy. So, private companies at the moment are trying to find new ways of collecting data, new bits of data that might be valuable that they can set on to consumers, the pharmaceutical industry or other people like that. Infect they even try to sell it to the public domain. Of course the public domain is been doing the genom because it feels that is so valuable to everybody it should be available to all. So infect the public domain can sit back and sort of wait for this little start-ups to try out new things and all the ones which don't work they gone get bust and the once that are going to be successful, well the public domain just will do all of it and give it away for free and that will go bust as well.

If I can just jump in with one little fact. I think Tim's shot that showed the number of sequences between 1999 and 2000 is not without significance because the bioinformatics industry looked very different in 2000 from 1999. And a lot of the so called first generation bioinformatics companies like Celera a really struggling now. And one of the reasons there are really struggling is because of a lot of information in the public domain. Another reason why they are really struggling is that they have sold their best databases or licensee them. And they are really can't get long term value on those databases until the drug companies they have licensed them to develop therapy and that could be 15 years from now, if at all. So, at some level you know Celera might remain a major player but they are not sure at this point that they remain a major player. They are quite confused about their business models.

Tim Hubbard: Yeah they are trying to become a drugs company now and it is quite a difficult thing being a drug company and selling data to other drug companies so you are competing with your customers. It does present some interesting thing you said about them having a revenue stream 15 years on. Well allegedly they are giving away their data without any clauses you know. Allegedly at least the public domain things that are being sold excess to Celera databases are that there is no reach for it. You see the data. They don't have any claims. So that is true if they do thing they gone get a revenue stream 15 years on then that means the private agreements which of course we don't have access to must contain clauses which assumed that they have getting there trying to get reach for. But I think it is pretty blink for private bioinfomatics because it is very difficult to grab something that is gone last.

James Love: I just like to ask Tim in looking at the problems you are looking at in terms of this kind of genetic research have you imagined alternatives to the existing intellectual property regime? And if so how are they work?

Tim Hubbard: So I mean I sort of got half way there by saying that I think people could do things in the public domain. The public domain certainly can do large scale things you don't need to invoke the need for private company. So, this question we have seen all this application in drugs where they claim they need the high prizes and the monopolies in order to get the research done. But suppose all the research was done in the public domain you wouldn't need that. You could convert drugs companies into efficient bodies to manufacture drugs. If you did all the research in the public domain you end up with molecules that have been tested that appear to be drugs. You put them out there and say these are drugs. Anybody can go and make them and then it could become superior manufacturing competition. Anybody can make it. They just have to have the right safety standards ect. But then you change the market completely and it might be a more efficient model.

James Love: I mean some people the way that they are look at the models they take a look at what the budget is in a country for medicine and then they sort of estimate what they thing the competitive prize would be and they look at the difference and say you know you could spend x-number of billion dollars in terms of that drug development thing and turn over to the (...generics) and you will be still better of. That is kind of how they kind of get that issue. A lot of people are not comfortably with the idea that you just wipe it out entirely the whole privatisation part of the development. I mean some people like Jerome Reichmen he has this model where you have a weaker right. You replace the strong right of patent for a weak right. And that is more of a compensation model I suppose in a non-exclusive bases. So if somebody uses your technology they have to give you some money for it. But the bases of it is not what is worth as a monopoly. Would that solve a lot of problems the genom to sort of move away from exclusive rights or is it different?

Tim Hubbard: I think it would help a lot but I don't think it would solve the problem of developing drugs that help people. Because it is quite clear the companies make decision about what they gone research on or manufacture purely on the bases on whether they gone sell enough to make it worthwhile. And if you take away the cost of development and just make it a manufacturing problem you can then have the development controlled on the research bases, central funds, maybe who managed do something so that you look for new developments, new drugs where it is clear that it's gone be a better drug or it's gone treat an area which hasn't got a drug at the moment. There is an awful lot of (...) to drugs out there which really aren't really better. And then you have a new marketing campaign by the companies to sell them against each other which isn't real helping anybody. One of those drugs would be enough. If you targeted which drugs got researched and then just make them available for manufacture then probably you could still sell with cost plus profit. And they probably would be still cheap enough then to be adopted anywhere in the world. So, even if it was a market in a third world country which wasn't very rich, it would still be a market that 's worth supplying. It is clear. India manages to make money out of selling this generics. Everything will be generic in this model.

Of course the model that you are talking about sounds all like the reward systems that people occasionally talk about as an alternative to propriety or not propriety but exclusive based payment systems. And they were actually used during the communistic period in the Soviet Union. And I have actually been looking for but haven't been able to find some really good data on you know how effective their actually were when the Soviet Union tried to use inventetaiv certificates for innovation. And from what I have heard a lot of failures had to do with bureaucracy and other problems that they had in the Soviet Union. But there might be some good data on what actually happens when you try and fill the system like that out there in Russian probably.

So, I think that you still have to have competition. I think you have to have that. Why has the genom project done so fast? Well, I have to admit you know Celera was a competitor and we couldn't stand him. But we have actually that sort of sounds like Celera did a good thing but infect there was an enormous amount of competition inside the academic steer here. Because we have this various labs around the world, all trying to do the largest chank of the genom, and infect Sanger was pissed of that white had did more then was in the draft. That we are doing more to finishing I have to say. So there is quite a lot of academic sort of pressure that can drive these things forward. Maybe we didn't have that in a totally planed economy.

James Love: We have seen one really troubling rewards type model that is the (crema sex) proposal on development of vaccines where they propose you put something like seven billion dollars and donates funds in to a reward for something develops like a vaccine against Malaria or something like that. It is a winner take all model. So, we have all this incremental research that comes of the...a lot of problems that Tim talked about are a big deal in the medical research. If you sort of shot of a sharing of information, the giving of papers, do this or that, you realy stay for research thing. And if people think that if they get the blocking patent on a seven billion dollar thing they might take...will go out hidden to sort of protect this kind of money. I mean it becomes a pretty vicious game at that thing. You have that case where big public investments then became (privatedist) to somebody who comes in at the very end. Another problem we didn't talk about is the problem of the incremental you know going back to Microsoft you know the software model it is the all embracing extend thing. They take somebody stuff and then they add that little bit at the end and they are trying proprieties that. It happens in the drug area quite a bit where the work will be done to a certain point in the public sector and then the big pharma guys will coming and they require some kind of last mile intellectual property rights of what type or another. And then they become essentially the owner of the whole value unit, the whole public improvement becomes their property. And you know that's why the JPL is such an interesting thing in the software because it is designed to go against the embracing extending. You guys even thought about JPL in your work and then rejected it. One idea would be in patent pools would you put him into effectively like a patent pool that required people that use the patents to make certain concessions to get in the patent pool. Sometimes like the JPL shot as long as you use strong property rights is a weapon against abusive propriety rights. I'm on the MSF working group from the docs from the collective disease. An issue for them to develop drugs is what they do with their inventions and I said what donate assume that you done want get any patents because you may be better of if you get a patent because then if (Smith Client Beach) get a little bit of improvement patent at least you got some bargaining power. They can't just sort of ignore you at that point. You going down a kind of spy role that you don't really wane be going on. But that is why I work quite interesting in sort of thinking about something other then the centuries all exclusive rights model.

Skuli: I have a question, just one tiny announcement. I have this one page handout from me on this Icelandic thing, so take it afterwards free of anything charged.
I learned so much from your discussions but I find them still a bit to abstract and I was wondering to the extent both for Hubbard and Love, but also for the others panellist on the panel namely, if we are dealing here with biomedical research where do the patience come in and where do the patience group come in? And somehow that is to say even lets stay within the western world, the industrial world but how does your models and your assumptions interact with real live people and patient groups ect, ect? In German there is this word Bodenkontakt, and somehow I feel as if there is no Bodenkontakt in what you are saying. You are floating above the ground and you are not getting your feed in the mood.

I want to provide a little bit of Bodenkontakt here. This is exactly what I was gone offer as fare alternative models. If we thing about the case of Patrick and Sharon Terry. This is a vary well known case at this point but there children suffered from a thing called Sudosentolasticom. It is a fairly obscure disease, one of only like two percent that are supposed to be single gene diseases. And what Patrick and Sharon did was travel around the world and collect information from all the people they could find who suffer from this disease, collect that information and then go to researchers and say: 'We have done the hard work. You help us sequence the gene and whatever work needs to be done to led to medical research and then we will negotiate who gets what share of the patent rights.', and then negotiated part of the patent rights. Now the patient advocacy group for this particular disease owns part of the patent rights for that disease along side the researches who sequenced the gene. Now, this is obviously a very special case because there is very, very few suffers from this but it can be expended actually quite likely. There are a lot of diseases that are organised around patient groups, that are organised the kind of advocacy people can do on the ground for that. And this actually in some bizarre way represents an alternative model for that but certainly not a very threatening one to the pharmaceutical industry. Because the issue is you know the pharmaceutical industry isn't gone develop a drug for that disease anyways because there is no market therefor.

I see the question here is that they don't wane make money they wane get control.

Right exactly. Well they also wane be able to charge license fee to anybody who wants to develop tests rather then having one particular agency involved with have exclusive rights with the tests.

Can I also add this last question. I think it is a interesting question in one of the reasons I'm avers personally to providing models though I'm totally for listening to other people providing models, is that when you perform a mapping exercise of drug development, intellectual property issues you find that that is a lot resistance to the intellectual property. a lot to that resistance comes from within the market. It is not just necessarily just external to market system. And I think we have already heard three examples of adding a public researches who want information in the public domain. Sypla, the Indian company that is providing AIDS drugs at low cost. Now, this isn't an active altruism. This is very, very (fruit) buisness that Sypla is performing. And then you have the patience groups who want to accelerate research but the want to have research in the first place but they want to have some control over the research that happens. Another set of resistance comes from native American populations for instance who been very resistant to the human genom diversity project. Because they really feel that the location of power centres as such that they won't have any control in the outcome of that research. So there is resistance to intellectual property because not they want research seeped up and to control it but because they really don' t thing that they want scientist to be researching their genes, whether those scientist are public or private. Si I think what is really intersting is being attentive to the stake holders and realising that there is a lot of resistance coming from often opposing stake holders.

It is actually a comment. Chris model you suggest for patiences, if you talk of something Malaria of course it would be equivalent to public owner ship of rights, of public funding of, because it works in very specific cases when there is a small patient group and it doesn't make sense for public funding to work but when there is a large patient body then it would be the same as public control of. So, it doesn't provide an incentive for private research in that sense. There is another model that has been written about for intellectual property for software and art, called the street performer protocol. I don't know if you familiar with that. I can't imagine that would work in bio-infomatics, but basically the idea is that the developer of the intellectual property has an incentive which is fixed, which is based on the cost of development and an undefined rate of return. It can be thousand percent or a million percent or what ever, that is not the point. But it is tied to the cost of development and not to the coast of production or the coast of a number of users.

James Love: In Australia they have develop a prize system on drugs. They have a fare more economic analyses where they take a look at the product and they figure out what benefits the product has. They ignore the cost all together. And they say to themselves given our budget what is the value of this product or another thing. And they work within a budget constrains. So the end of the day all they allocate is what there budget is. Then they just allocated the proposition amongst products according how much value they think they have on medical grounds. The problem for a poor country is you have to have a lot of experts. It cost a lot of money to sort of do that right, but it is a sort of an elegant procedure. If you wanted to fund RMD it thought it would be interested to in sort of the prize model I refered early on. So if you could sort of say o.k. for an experiment we put like you know five hundred million dollars or what ever. What Tim said about this the researches are going to the public domain and people could apply to sell stuff to the public domain. But then you would have a panel people who sort of allocate things based upon what ever criteria's. The problem with something like that if you created something like that maybe nobody wanted to put it in the public domain for free. I don't know. There could be kind of one intended consequences that somebody things so. I think part of what you need to do this big privatisation initiative at the university level in the United States and now in Europe toward pushing for more propriety models intellectual property. I think there is recognition that is impeded a lot of academic research and sort of shrinking the public domains. So people looking for like ways, vehicles like the snip thing or things like that, but to kind of more in the public funded model or mixture of public and private, the kind of protected public domain or even find different systems I'm not sure. Looking at the cost of drug development is going to be a problematic for people because it is the clinical trials is easy to do it that way. That is actually fairly simple but the preclinical is really totally it is so risky. It is so opportunistic what happens there. It is kind of hard to attribute where the value came from for the clinical side. There is so much capturing others peoples ideas and then kind of making this missing links and stuff .

But that's no different from normal research science which is also incredibly risky.

James Love: But I mean I wouldn't want to look at the investments you made on the preclinical side is the key towards saying what you want send the people in that market, I'm not sure. I mean unless it was like protecting investment. If you were trying to protect investment then it is the right thing to look at.

The only thing you protect in a scientific world is you want to be able to get grants again, so you won't be out of publish, you want to get credits having been smile and discover things and that is enough.

That is sort of your street performer protocol as well. You... and then you give it away afterwards.

Exactly. Except actually you normally have to bug get new grants ...it is very similar actually.

Skuli: I was wondering I found what Hubbard said very sort of optimistic from the Icelandic resistance side namely the blip for the future of bioinformatic companies. And I was wondering just as patience were not spoken of proteins were not really spoken of your either and if you read for instance recently Richard Loenton in the New York review books then you would say what a lot of other people had said that now that the hangover period is going on after the human genom drinking party that somehow genes want to really that much for you and now we have to look at a protein but probably with some difference elling tricks because the old once were used for the genes. But what will happen to your economic model once people shifted attention ever more ever more to proteins and saying that this gene fixationis just an illusion. And under the gene as an concept will crumble under the weight this data base as also this theoretical work at percent would allude to like evident (...) and others namely that a gene want to (...) the human genom project. What would then happen to your models?

Tim Hubbard: So I think that the only real areas that you got to worry about is large datasets that are important. So the genom and the genes is a large dataset. No once found the good way to deal with a lots and lots of proteins. I said before it become clear the value of large datasets, so if anybody does find a good way of dealing with proteins on large scale we will do it in the public domain anyway. For everything else it is probably gone go on just as its been going on for the last 15 years where individual researchers work on individuals proteins and trying to work out how they work. So, the game is changed in some ways it becomes small scale again. Unless somebody finds a smart way of doing of large scale and we know how to do with large scale. So, is that answer the question? I don't think it is a serious problem. If there was another Celera like monopoly threat then I'll be worried but we think there isn't a similar monopoly threat in the protein. You have to be careful about this own business. It is all just marketing spin.

Skuli: But I was thinking more about if this bioinformatic companies now are in trouble in some sense and if then the proteins...

Tim Hubbard: Proteins aren't gone save them....

Skuli: No, I was rather thinking of the Titanic effect. How quickly will they go down and at what rate and will your models include that?

Tim Hubbard: There is still certain amount of data. There is a lot of genoms to sequence. There are companies that are collecting data on a range of different organism and selling them later on. And those aren't efficiently critical that it is necessary to do them all in the public domain. There is an infinite amount of sequencing, infinite amount of ictes to collect. So, there is room for people to do things, so there are more in demand by private companies, that are more valuable for them but aren't critical for public domain research. Because public domain research hasn't gone yet into the pure drugs area. Now it if would got involved then we would have to tackle this as well.

[transcript: Katja Pratschke]

Creative Commons License
All original works on this website unless otherwise noted are
copyright protected and licensed under the
Creative Commons Attribution-ShareAlike License Germany.