GEOFF MULGAN: Jim Watson, as one of the people most responsible for the development of modern genetics, how do you think history will rank the importance of the Human Genome Project?
JAMES D WATSON: Having the human instruction book is a seminal moment in history. We can read the instructions [the genotype] which lead to the development and functioning of our bodies [the phenotype].
STEVEN ROSE: I am uneasy about this metaphor of the instruction book. It doesn?t capture the way genes interact with each other or with other components of the cell. The implication of the instruction metaphor is that the rest of the cell is simply there to carry out the demands of the genes. This misses the dynamic in which control of cellular processes is vested in the cell as a whole, not just in one of its parts.
JW: So what else is the instruction book?
SR: It’s not that something else is the instruction book.
JW: You underestimate what’s been achieved. When we began, in 1944, no one dreamed we’d get this far. Now we have the code script: it’s a language written in four letters, and we have the book.
ROBERT PLOMIN: As a user, I think it will be very valuable at a basic science level. Of course there are many questions about how we interpret these lists of letters, but it’s a momentous occasion.
STEVE JONES: This is one of those rare things: a breakthrough. But what we do with it is not particularly a question for scientists.
NANCY CARTWRIGHT: I agree. We can’t progress without knowing what the science is, so you scientists have to be at the table. But scientists by themselves aren’t in a good position to judge how science will be handled when it gets out into society. For that we need the help of social scientists.
GM: Who should own this body of knowledge? Without a patent system which allows organisations to benefit from their own inventions, there would be little incentive to do scientific research for economic gain. But the argument is that the genome is public property, it needs a different structure of ownership.
SR: The argument against patenting the genome is that it exists in nature—it is not an invention.
PETER GOODFELLOW: I am in favour of the patenting system but I find it very confusing that, for example, if I create a new variety of plant I am not allowed to patent it, whereas if I invent a new paperclip, I can.
SJ: There is a consensus—with some noisy dissenters—that there should be open access to this knowledge for the scientific community. Most scientists are shocked to find that it’s being hijacked. As Steven Rose says, the distinction is between an invention—which should be patentable—and a discovery—which should not. Some people who want to patent DNA are evading this distinction; they even try to patent sequences of DNA without knowing what they do.
PG: There are two issues. The first is: do we have to change the law? The second is: is the law we have being applied? Successful patents must pass three tests: the invention must be novel, useful and non-obvious. These things can be shown in court.
SJ: Speaking as someone who has had patents infringed, may I just say that testing the law is expensive. Keeping a patent alive can cost hundreds of thousands of dollars a year. This means that money will win in the end. And that is what is happening.
SR: The US patent office has been making the running on this. The US authorities are allowing too many things to be patented, and many Europeans have been dragged in directions they didn’t want to go.
JW: Because of the political implications of the genome, I would have been happier with a system of obligatory licensing rather than patenting. So, for example, whoever discovered the breast cancer gene would get a royalty licence, but not a monopoly on the right to develop a breast cancer test. Likewise, licensing could apply to all the genes which are targets for drug discovery, giving a 1 or 2 per cent royalty—not affecting the price of drugs much, but providing some reward for the discovery. In retrospect, I regret not raising the patent issue at the start, but I suspect that no one would have listened until we got into the mess we’re in now. I don’t think we are going to get licensing. We will get a mixed bag of patents. People are getting patents without having done a single experiment; but patents should have to be non-obvious. In the US, Harold Varmus and Francis Collins have gone several times to the patent office to argue for a more restrictive approach to patenting, but they have been disregarded. That offends me.
PG: But the courts can sort out a misuse of patents.
JW: The trouble is that what is obvious to a judge is not obvious to a DNA technician. There is a lot of apprehension around. Some people who want to use DNA knowledge are put off by the patent thing.
GM: Let’s move on to genetic medicine. So far, relatively few techniques or products have come on to the market. But are we on the brink of a flood? And what are the implications for our health services?
PG: Healthcare is a very regulated environment. To get a new medicine on the market we are talking about cycle times of ten, 15, or 20 years, so the euphoria several years ago about DNA cloning [which allowed the expression of proteins in a test tube which could be used as drugs] is only now producing results. It will be at least ten years before medicine is significantly affected by discoveries taking place now.
SJ: Do you include diagnostic medicine in that?
PG: For the monogenic diseases [diseases caused by inheriting a single gene defect] there is already an enormous impact. You can identify individuals who are at risk of passing on such destinies, or identify people who have inherited such destinies. In Britain, but not in the US, this is regarded as a boon, because it enables families to decide whether or not they want to have another child afflicted by a particular genetic disease.
JW: Yesterday I was talking to Middlesex medical students. I asked whether any of them had been screened for cystic fibrosis. No one put up a hand. I don’t think that what we have learned so far is having an effect on the incidence of cystic fibrosis.
PG: If you look at families with a history of cystic fibrosis, screening is happening, and every family in Britain which is at risk of Huntington’s disease has been offered a test.
SJ: They’ve mostly turned it down. But people are, actually, quite rational about this. On something like Huntington’s, which is a late onset illness about which nothing can be done, what is the point of knowing? But on something like colon cancer, which if diagnosed early enough can greatly increase your chances of survival, most people accept the test.
PG: Looking beyond the monogenic diseases to the wider morbidity and mortality issues, you start to get into a more difficult scientific and cultural area. If you go to your doctor and he takes your blood pressure and it is high, or if he takes a sample of your blood and it shows that you have very high levels of cholesterol, then you are at greatly increased risk from a variety of unpleasant diseases. Genetic testing for these diseases will be rather similar—the result is not your destiny, but it is indicating a particular risk.
RP: But this is where the genome project will be most valuable. We didn’t need the gene sequence to find most monogenic disorders. But it’s with the complex disorders, involving many genes with small effect, that the gene map and especially the SNP map [single nucleotide polymorphisms, which reveal the DNA variations among people] will be most valuable. But first you need the sequence, before you can start talking about the variations.
GM: Some people predict that soon we will all be screened in detail for our genetic predispositions, providing employers, insurance companies and so on, with far greater potential knowledge about our fate.
SJ: We can already predict very accurately people’s fate from one piece of information relative to their birth: their postcode. There is an 11-year difference in longevity between babies born in the richest and the poorest postcodes in Britain. We know that and we cope with it. It’s unfair, but we cope with it. And I don’t actually see that multigenic genetics is going to make much difference. Take heart disease. You can make a better prediction about the likelihood of heart disease from a cholesterol test than from a huge screening for the many genes which have been found to predispose for heart disease. So it may be that this great map, telling us all our predispositions from our DNA, may be a bit too optimistic.
SR: Yes, people have stopped being so optimistic about gene therapy [the process of implanting a new gene sequence to replace a faulty one]. Talk now is much more about pharmacogenetics—looking at how genes determine your response to different drugs. Take blood pressure: at least three different classes of drugs can be used to treat hypertension, and we don’t know which one is going to be effective on which person. The argument is that if you know the polymorphism [the genetic variation between people], you can predict which of those drugs is the most appropriate.
PG: This comes down to whether we will be able to use genetic information to get a better definition of disease, and thus a better treatment. My fear is that, in many cases, it might just be too complex.
RP: I think that the scepticism which some of you have expressed is unjustified. In most cases of illness or disability, we are talking about lots of genes with very small effects. So I agree that it’s going to be very difficult to pin down causes—there won’t necessarily be many new drugs. But in terms of diagnosis and prevention, there are potentially big and hopeful consequences from the new knowledge. Take reading disability, or many other developmental disorders. If you wait until children have such problems and their worlds start falling apart, it’s harder to do anything about it. This applies to obesity and alcoholism, too.
NC: If you could detect a dyslexic child in the womb, what would you do that we don’t do now?
RP: The first point, which we didn’t know ten years ago, is that dyslexia is highly heritable. If you didn’t know that, you might think it’s something the parents had done, and you’d be on the wrong track.
SR: You don’t need the sequence to know that.
RP: But we can do better than that. There is a specific chromosomal region linked to dyslexia in half a dozen studies. We don’t have the gene itself yet, but it is one of the linkages which seems to be holding up best. There will probably be lots of such genes—many with very small effects. But, assuming you could get enough of those genes to make a prediction, it would be useful to be able to predict that a child is going to be reading-disabled if we allow the normal course of events to take place.
SR: I would simply reduce class sizes.
SJ: But Robert has a point. Take the alpha one antitrypsin enzyme deficiency in the lung. If you have a rare form of this and you smoke, you will die horribly at 40. In Sweden, now, they screen all children for this at birth, and the incidence of teenage smoking among those at risk is less than half the rate of those who do not carry the gene. So we’re being too self-critical to say that the new knowledge will have little effect.
GM: So much for medicine. But what can the new genetic science tell us about genetic predispositions towards certain kinds of behaviour?
SJ: I think most of us around this table have set exam questions on that question and, with one or two exceptions, it is almost impossible to answer. Genetics is often used to reinforce existing prejudices. At the moment we live in a rather liberal environment, so genetics is used as a sort of forgiveness for certain behaviour: “he was born that way.” But it could swing back the other way. The most dangerous word in genetics is the word “for.” The gene for this, the gene for that. It has a very limited meaning. It has a meaning, but it means much less than the word “for,” in general parlance, because genes and environments work in consort. You can cure many genetic diseases by changing the environment.
RP: We may not find specific genes responsible for specific behaviours, but over the last 20 years we’ve gone from thinking that, say, most mental illness or reading disability or autism is environmental, to seeing them as genetic. I think that’s a dramatic development. Quantitative genetics consistently points to substantial genetic influence—so much so that I think the way forward in research is not to continue demonstrating heritability, but to roll up our sleeves and find some of the genes responsible for this genetic influence. We are not looking for a single gene which is wholly responsible for a disorder or a behavioural predisposition. We may be looking for many genes which have very small effects—perhaps even less than 1 per cent.
JW: I think that 1 per cent is unlikely. I think we’re looking for effects of 10 per cent and 20 per cent.
RP: Not for mental disorders. My bet is that, as a rule, we won’t find any genes with effects as significant as 1 per cent. But there may be single genes which in a very small number of people have a significant effect.
GM: So, how far are we from being able to shape the genetic inheritance of a child, not just screening out the various negatives?
SR: We are not going to see designer babies or breeding for higher intelligence or for baseball skills or whatever. That is beyond the competence of science, and it’s also beyond the power of genes.
PG: But we can discover the sex of a child very early. And we can check for monogenic diseases.
GM: So the rather unpleasant prospect is: ever more sophisticated negative screening, plus abortions.
JW: Why unpleasant? It’s better than having a child suffering from a serious genetic disease.
PG: Or a child you don’t want. About 95 per cent of all abortions are carried out for purely social reasons.
SJ: There are tens of millions of abortions worldwide each year, and if we accept that, then why not use our genetic knowledge?
JW: I think it’s wonderful that we can diagnose a Down’s syndrome child early and offer the choice of aborting or not.
PG: If you allow people to have abortions for social reasons, how can other reasons, based on genetics, be judged any less acceptable?
JW: I think very few women will do it, but if they want to, I wouldn’t prevent them.
GM: You would set no limits at all?
JW: No limits. Parents should decide.
SJ: In Britain the general public is more willing to accept selective termination than professional geneticists are. One survey found that something like a quarter of British parents would accept a termination for a foetus which has two missing fingers.
NC: The views around this table about what we should do about aborting foetuses with blue eyes or Down’s syndrome are irrelevant. What we need is a mechanism for taking these decisions in a serious and informed way. At present, the decisions are taken in a haphazard way.
JW: Do you want a committee of wise women like yourself telling other women what to do? You want a pseudo-consensus which, in practice, takes the decision away from individual parents.
NC: I want a serious study and a serious public discussion.
JW: You mean: more social science crap.
NC: More social science information. I think it is too bad that biologists find social science so hard that they fall back on taking decisions in an uninformed manner, based on loose, abstract, moral ideals, without a clue about what would really happen if a particular proposal were to be adopted.
GM: As a summary, would it be fair to say that what has been learnt in the last few decades about genetics has confirmed powerful views on difference between sexes, destroyed any notions of difference between the races, and said almost nothing about class?
RP: I am a behavioural geneticist, which means that I focus on individual differences within groups. We don’t have very good tools for determining the differences between group averages—even between the genders. You can say it’s the X chromosome, but it’s very difficult to pin down what is the genetic source of difference between genders. Between class and race it’s very difficult, too. Even if you had a bit of DNA in a different frequency in two populations, how do you prove that that difference is responsible for the physical or behavioural differences you observe? By contrast, we have very powerful approaches to understanding individual differences within groups.
SJ: There is a tendency to rediscover the blindingly obvious in a lot of behaviour genetics. The reason that men are more violent than women, which is observably true, has something to do with the fact that men have a Y chromosome and lots of testosterone. That doesn’t seem to me a shocking or illiberal thing to say. It is a genetic statement of difference between sexes. But it’s also obvious that the levels of violence within societies and between societies have got nothing to do with genes. We are much less likely to find biological differences between races, simply because the genetic distances among them are so small.
SR: There is a tension here between the evolutionary psychologists and the behaviour geneticists. Evolutionary psychologists say that there are human universals which were laid down in the pleistocene epoch. They claim, for example, to have established the genetic basis of behavioural differences between the sexes—the fact that men prefer younger women, that younger women prefer older and richer men, and so on.
JW: I don’t see what evolutionary psychology has to do to with it. We know you don’t like it.
SJ: Evolutionary psychologists have a science of similarity, of human universals. But as geneticists we are interested in the science of difference. Why is human A not like human B? But there is an interesting problem here which is seldom acknowledged. Evolution is a comparative science, but when you come to things which are human, nearly everything about being human—as far as we can tell—is unique.
GM: When people see teenage delinquency or violent crime, they increasingly look for a genetic or evolutionary explanation rather than a social or historical one. Are we saying that they will be disappointed 99 per cent of the time?
RP: No, I think people are right to have a greater awareness of genetic causes, but I am still amazed at how many people, even educated people, do not apply it to their own lives. For example, several of my yuppie friends are (a) infertile, (b) thinking about adoption, and yet (c) not thinking about heritability. At some level they still cling to the idea that tender loving care is the only factor in raising kids. But some of these friends, who have adopted, have been disappointed to find their child is of only average intelligence—which in these families means almost retarded. Genetic and evolutionary explanations are accepted in theory at species level, but at the level of how people live, they have not been taken on board.
GM: But, given the normal distributions of characteristics of any kind, if you become a parent, whether biological or not, you accept a randomness about what your child will be like.
JW: No, I think we generally assume that more intelligent parents have more intelligent children.
GM: But with wide variations.
JW: If you marry a stupid girl your children may disappoint you.
NC: I want to go back to this question of the evolutionary psychologists. Some of you seemed sceptical.
JW: I’m not, I am one.
NC: But I want to ask you, Robert, the behaviour geneticist, what kinds of things your studies show us. Where are your success stories?
RP: My favourite story is autism. In the late 1970s there was a review by geneticists which said: “Well, here’s one disorder which doesn’t seem to show any evidence of genetic influence,” but there hadn’t actually been any good studies at that point. Subsequently, Michael Rutter compared identical twins [who are like clones], to non-identical twins [who only share half of their genes]. He found that the identical twins were about 60 per cent concordant; that is, if one is autistic, the chance that the other is autistic is about 60 per cent. By contrast, the concordance for non-identical twins was only about 5 per cent.
NC: But autism is easier to describe in these terms than, say, a behavioural characteristic such as aggression.
RP: The truth is we have found nothing which is not significantly heritable, including intelligence, which most geneticists say is about 50 per cent heritable.
PG: Everything has to do with both DNA and the environment, unless you believe in God.
RP: Social scientists didn’t believe that in the 1960s.
JW: There still seem to be people around who believe that it’s all due to the environment. I remember, a few years ago, you, Steven Rose, saying that most of schizophrenia is due to rampant capitalism.
SR: That’s absolutely not what I said.
JW: Everyone applauded because you were pandering to their prejudices.
SR: I have never said anything as nonsensical as that in my whole life. What I have said is that the diagnosis of schizophrenia is complicated and that the genetic arguments so far have not held up. Some people at Robert’s Institute of Psychiatry, such as Robin Murray, argue that schizophrenia is disappearing.
SJ: What about velocardial facial syndrome [a deletion of part of a chromosome, which gives a sweep of symptoms that can look like paranoid schizophrenia]? How can you deny that genes are involved?
SR: Who is denying that genes are involved? I’m not.
JW: You indicated that they were quite unimportant. Incidentally, what is the concordance in schizophrenia in identical twins?
RP: It’s about 50 per cent.
GM: Some of you may have read Ronald Dworkin on how genetic science dislocates the boundary between chance and choice, which is the spine of our morality. Is there a problem here?
SJ: I have faith in the public’s ability to get bored. They are already getting bored: “Genetics, well, isn?t that a bit 20th century?” It will be assimilated into the corpus of knowledge and experience in the way that medicine, or cars, or television, were. It won’t make much difference to the way we live, or to our morals.
GM: Or to our commonsense ideas about personal responsibility for behaviour?
JW: The main moral dilemmas are to do with the additional knowledge about disease and the hard choices with which our knowledge will present us.
SR: The most sensitive areas are behaviour genetics and the whole terrain of so-called smart drugs—drugs to improve cognitive performance. Such drugs could have important consequences, and we need to be proactive in thinking about them. But I agree with Steve Jones that our moral framework won?t be fundamentally changed.
GM: So the arguments won’t be different in kind from 30 years ago, when we might have said that social factors explain behaviours and crimes, and excuse them, too.
SR: No I don?t think it will be any different from saying that you are a violent criminal because of the way you were brought up, or to say you are a violent criminal because of your genes. It doesn't alter the question about responsibility.
PG: I have one caveat. Sometimes in history there is a rapid rate of change. We may be undergoing a rapid rate of change in our knowledge base now, and that may be hard to assimilate. But in the end we will use things if they are useful—and not use them if they are not. Getting to that point may be a bit bumpy.
RP: At a science level I can’t imagine anyone disagreeing that having the genome sequence is going to be a very useful tool. But at a practical level, it comes down to how much of these complex traits and disorders we can actually predict from our genetic knowledge. A lot will come down to cost-benefit ratios. We have been doing genetic screening for decades for single-gene disorders such as PKU [phenylketonuria, which causes severe mental retardation], because in that case there is a simple intervention which works pretty well. Now there is a lot of new information which says that common disorders are not simply the result of one gene—but hundreds of genes, each with small effects. As Peter said, it’s going to come down to: is it useful? Will our genetic knowledge allow us to change the environment in some small way—changing a woman’s diet during pregnancy, say—which makes a difference for children with a certain genetic risk?
GM: There are two specific concerns about this genetic knowledge, relating to health systems. One is that the incentives for pooling risks may break down because more people have reasons to opt out, knowing that they are healthy. Second, if you have the knowledge about individual predispositions, is it legitimate to expect greater responsibility on the part of an individual to reduce a risk?
SJ: Insurance companies in the US hate the new knowledge, because insurance of any kind depends on the company having more knowledge than the people it is insuring. In fact, the knowledge damages both parties, because only those at risk buy insurance; those not at risk don’t buy insurance, so the whole thing collapses. And it’s noticeable in the US that the insurance companies are combining into larger and larger consortia in order to diffuse risk. So you end up with something like a national health service. The question is: do you want a police force which protects everybody, or do you want a set of security guards who only protect those who pay for them? Genetics allows us to move towards a set of security guards: because I know I am not going to get cystic fibrosis, I am not going to get lung cancer, so I don’t have to bother. But the experience is that a universal police force is always better.
GM: Nancy, are we moving to a happy new collectivism of pooled risks?
NC: No. But first I want to argue very strongly against the tendency to pronounce casual opinions about social implications—about what should be done, or about who should decide various matters. All of you sitting round this table have some pretty well-founded guesses about some of the effects of certain genes. But I am certain that no one around this table has the kind of knowledge it takes to make well-founded guesses about the social implications. Here we need to rely on our social scientific knowledge about our own legal and social institutions. Of course social science has many gaps. But social scientists know a lot more about these things than we do.
PG: Tell us what study you would commission which you feel would help us resolve your difficulties.
NC: I don’t know. I don’t know enough about our legal structures, our social institutions, our families, to think about designing a study, any more than I know how to design an experiment identifying genes. That’s why I want these questions discussed in detail by people who do know what there is to be known.
GM: Looking ahead ten or 20 years, do you think the gap between the favourably genetically constructed and the less favourable genetically constructed is going to widen or narrow?
JW: If genetics is used properly, the gap will narrow. There will be fewer children born with very serious problems. The problem is that instead of society telling people to use genetics, we’re telling them not to use it, we’re making them feel that they’re doing something evil, that they’re playing God. I don’t think you in Britain are doing any better than we are in the US—there is little screening except for Down’s syndrome.
SR: We would make more difference to the well-being of the British population if we dealt with the 20 per cent of children who are living in poverty than if we invested in huge screening programmes, desirable as those programmes might be. Our main social problems are not going to be affected by genetics.
PG: I agree with Steven Rose. I worked for 13 years for the Imperial Cancer Research Fund, and the biggest effect you can have on cancer in our society is to ban smoking. But we choose not to. So the social and political issues remain more important than the new technologies, in the short term. In the long term, I am an optimist. There will be many discoveries under the heading of the new biology, the new genetic medicine, which will give us tools to reduce human suffering.
RP: Abolishing poverty and getting to the genetic root of many developmental problems are not mutually exclusive. I would like to identify some more specific genes, and understand how they work on the brain and on behaviour. I know that’s going to take a very long time, but we can do it.
GM: Steve Jones, is this going to widen the gaps between people’s life chances or narrow them?
SJ: Neither. Genetics, historically, was the science of exceptions; it dealt with rare, inborn, diseases which couldn’t be treated, and that’s what it will be in ten years from now.
NC: Well, consistent with my view that one should have systematic knowledge before coming to an opinion on a serious matter, I don’t have an opinion.
JW: Where did you go to school? You don’t have an opinion on anything.
GM: I think we better stop there. Thank you.
