Philip Ball is a science writer, and ex-Physics Editor at Nature magazine. He thinks the idea that there is 'a gene for every trait' is very misleading, and that biologists are still a long way off properly understanding the complex ways in which genes interact with each other and the environment.

Will the human genome project provide the information needed to find treatment for disease?
I think that really the genome project represents a beginning rather than an end point. It undoubtedly contains a tremendous amount of useful information, the only problem is that it's going to be a long time before we know how to understand and make use of that information. What has filtered through to the public consciousness is the idea that for every characteristic or trait or disease there is a single gene that controls or determines that outcome - we've even heard talk about 'the gene for homosexuality', 'the gene for intelligence', perhaps even 'the gene for criminality'. So a perception has developed that the genome project has provided something like a lookup table for all diseases, so that all one needs to do in order to develop a cure is to look in the genome for the gene responsible for that disease and then that will suggest a way in which the behaviour of the gene can be modified to develop a cure.

The problem is there are a few well known high profile diseases that do have a single gene basis, for example cystic fibrosis and haemophilia, but most single gene diseases are quite rare and most important diseases arise from the interaction between perhaps dozens of genes, in ways that we don't yet in many cases understand. I think this is illustrated by the fact that if we knock out the genes in the human genome one by one, we find that something like 40% of the genes can probably be knocked out individually without any discernible physiological response, at least in the short term - somehow when that happens the other genes are able to collaborate and conspire together to make up for the missing gene. This is an illustration of the way that genes act together, so that it's not always straightforward to predict what the consequences will be of knocking out or modifying a single gene.

Has the media therefore raised expectations too high?
There's a famous quote of Einstein that everything should be explained in terms that are as simple as possible but no simpler, and I think there's a danger in the case of genomics that things have gone slightly beyond that line, so that the public have received is this idea of genetic determinism and a linear idea of a single gene leading to a single end result. I think that that while that's partly a consequence of the understandable simplifications that the scientists are trying to make, I think it's probably also a bit of an outcome of our tendency to hope for a quick fix - if we can pin a particular disease on a single gene we feel we've nailed the problem, we know what to do, we can cure it. The natural desire to see that happen I think has encouraged people to want to interpret genetics in this way, the sad truth is that in very few cases is it really that simple.

Some have used the metaphor of the Book of Life to describe the human genome. Do you think this is a helpful way to think of it?
The scientists who've been working on genome sequences have understandably needed some kind of metaphor to try to convey what it is they are doing, what it is they are looking at. The one that they've hit on is to say once we've sequenced the whole of the human genome we've essentially read the Book of Life. To me that seems a little bit like saying that once we've read the English dictionary we have assimilated all the works of Shakespeare. In a sense it's true that if you work through the dictionary from cover to cover than you will in some sense have covered the works of Shakespeare - in fact not quite so because there are words in Shakespeare that won't appear in the English dictionary, and the same is true for the genome. But to me that's the sense in which the genome is the Book of Life and quite clearly we learn absolutely nothing about what happens in Hamlet or King Lear from reading the dictionary. So it seems to me to be a more accurate metaphor to talk about the genome as a dictionary. In fact it's even worse than that, because it's in this case a dictionary that's 95% meaningless junk where many of the words appear again and again, endlessly repeated over pages and pages, where the words are listed without their meaning being attached, where many of those words have no discernible meaning and where they are certainly not in alphabetical order.

How do genes work within a cell?
Some people think of a cell as a linear string of genes switching on and off one at a time. But with new genetic technologies it is now possible to obtain a snapshot of a cell and all the genes that are active at any one instant. This picture shows us just how complex the behaviour of the cell is, we find there are thousands of genes doing something or other at any one time. Making sense of that picture is a very big challenge, not least because often the genes that show up most prominently aren't necessarily the ones we are interested in, they're genes that have to be active in just about anything we do, they're so-called 'housekeeping genes'. But often the activity that we might be interested in is going on in one little corner of that snapshot. In general, interactions of the genes give rise to robust characteristics of the cell. In a sense the cell is designed in a similar way to the way engineers design systems to produce the same end result in a set of very different circumstances. The way genes do this is to operate together in networks, using principles very similar to those engineers use, principles such as feedback, back-up systems, error correction and so forth. It's the understanding of how these sorts of principles are used to provide the cell with its robustness that's really going to be needed in order to make sense of how genes give rise to the functions of life.

Scientists have been working to create a detailed picture of the human genome. Do you think this is the correct approach?
If I were to try to give a picture of the traditional idea that has developed amongst molecular biologists over the past decade or so of how genes work to control life, I'd again draw an analogy with Hamlet and trying to describe what goes on in the play. Instead of asking, 'What goes on in the cell?', we ask, 'What goes on in Hamlet?'. The traditional explanation would go something like, 'Well, two soldiers come on stage and they have a chat and then another one comes along and they talk some more and then a ghost appears and one of the soldiers says a bit like, "doesn't the ghost look a bit like the dead king?"', and so on. This step-by-step account isn't really what we're looking for. What we want to know in answer to the question, 'What goes on in Hamlet?' is, for example, 'Well, a young man struggles with his lack of resolve'. Those sorts of questions aren't necessarily the ones that can be answered by a blow-by-blow account, in fact sometimes that sort of account obscures those kind of answers. This is a leap that needs to be made in modes of thinking - it's really a different kind of answer that one is looking for, an account of the general characteristics of the cell that on the whole remain pretty much invariant despite what's thrown at it, how the broad themes emerge from the specific details.

What could geneticists learn from other fields of science?
Other fields of science can potentially offer some useful tools and concepts. For example, in physics the idea that one can find robust kinds of behaviour emerging from messy specifics is one that is quite standard. If you think of a glass of water, all those molecules are moving about in a crazy frenzy that is a completely different crazy frenzy from a different glass of water, yet both glasses will freeze to ice at 0 centigrade - that's a general property, something that emerges collectively from the way that water molecules interact with one another. We can expect to see at least some principles that are somewhat similar to this emerging from the interactions between genes. There are tools that have been developed within physics to handle big systems of many interacting components and to find out the modes of behaviour that emerge from those interactions. I think that it's likely that some of these tools may be useful in decoding what genetics really tells us about the way life works. There is some indication that this new way of thinking is already beginning to emerge within the biology community and that there is some recognition of the need to import ideas from other sciences. There are some molecular biologists who are trying to build models of the cell at the kind of systems level, I guess you could say, so that one tries to look at collections of genes and gene products and develop models of how they operate collectively to bring about some particular process. These models often involve quite sophisticated computational techniques or mathematical techniques and that is certainly one area in which physical scientists can contribute their expertise.

So, will the mapping of the genome transform medicine?
What the human genome gives us is a very useful medical database. If that doesn't sound quite as exciting as some of the claims that have been made for it, so be it. I think that even when we get to the stage where we understand what every single gene does - and we're very, very far from that stage - it will still simply be a database. Until we have a better understanding of the way that genes speak to one another, affect one another's behaviour and collectively act together we're going to be a long way from having a clear picture of how to attack very many genetically based diseases.

  < previous   next > Page 2 of 7