Peter McGuffin, who studies the role of genetics in the shared behaviour traits of families, continues a series in which leading academics discuss their ground-breaking research

Hardly a week goes by without a headline announcing that scientists have discovered the "gene for" yet another disease or trait of personality: genes "for" criminality, homosexuality, feminine intuition and even bad luck. (Embarrassingly, I was fingered, inaccurately, in reports earlier this year for the last discovery.) We have heard too, of genes that make people vulnerable to mental illnesses, rare ones such as Huntington's disease, others more common such as Alzheimer's disease and depression. Such reports excite not just interest but also controversy and, sometimes, hostility. Some exasperated biologists argue, correctly, that genes are not "for" anything other than reproducing themselves. And some social scientists point to the dangers of genetic determinism and the potential misuses of the idea that deviant behaviour might be innate. Researchers like me, interested in the genetics of normal and abnormal behaviour, have been the target of harsh criticism, portrayed as anything from wacky boffins to evil eugenicists.

What is it I do? In my work in behaviour genetics I try to find explanations for a phenomenon that everybody recognises - that many human characteristics run in families. The questions is: to what extent are behaviours and traits genetic and to what degree are they influenced by the family's shared circumstances and environment?

There are classic natural experiments that help in discovering whether a trait clusters in families because of shared genes, shared environment or a combination of the two. By comparing biological twins who have been adopted and reared separately with those brought up in the same family by natural parents we can learn much about the effects of nature and nurture.

³ÉÈËVRÊÓƵ

Identical, or monozygotic (MZ), twins share all of their genes. Non-identical, or dizygotic (DZ), twins share 50 per cent. If we assume that the degree of sharing of the environment is roughly the same in both types of twins, then greater similarities in MZ than in DZ pairs should reflect a genetic effect. Studies of twins have been used to examine the genetics of physical diseases such as diabetes and multiple sclerosis as well as those of mental illnesses such as schizophrenia. Indeed, the combination of twin data and evidence from adoption studies, showing a high risk of schizophrenia in the biological but not the adoptive relatives of patients, suggests that schizophrenia is substantially a genetic disorder, probably involving abnormalities in brain chemistry.

We can use such research to investigate the interplay between genes and environment. There is much evidence to support the commonsensical idea that depressive disorders are associated with unpleasant happenings. A few years ago, with colleagues at the Institute of Psychiatry in London, I studied depressed patients and their close relatives. We found, somewhat to our surprise, that not only did depression run in the families, but so also did unpleasant life events. Anita Thapar followed this up in a study of Welsh twins. She discovered that what may be inherited along with vulnerability to depression is a gloomy, threat-sensitive outlook, which may, in turn lead to unhappy life events, such as divorce. This is being explored further in a study of depressed adults and their siblings.

³ÉÈËVRÊÓƵ

When these studies attracted media attention earlier this year, we were billed as having discovered "the gene for bad luck". This is not true: we were not studying bad luck, and the sorts of traits we are interested in are likely to have very complicated inheritance involving the actions of several genes. Moreover, we had not isolated a single gene but had simply presented statistical evidence of genetic influences.

Yet one of the most exciting prospects in modern behaviour genetics is that there are now methods for going beyond abstract statistical inferences and really locating and identifying genes. This is made possible by the huge advances that have taken place in mapping the human genome, the 23 pairs of chromosomes that carry our DNA. There are now genetic markers closely spaced throughout the genome that can be used as reference points to track down genes influencing diseases or other traits. Rare single-gene disorders have succumbed to this approach with astonishing rapidity.

In neuropsychiatry, the Huntington's disease gene and genes causing early onset forms of Alzheimer's disease have been identified. By adapting the strategy, it is possible to move on to tackling more common, more complex disorders that involve multiple genes plus environmental factors. Mike Owen and I are completing a look at several hundred DNA markers in a study of 200 sibling pairs, in which both siblings are affected by schizophrenia. If the siblings share any of the marker types significantly more than expected by chance, this will give the location of the genes causing their joint susceptibility to schizophrenia.

An alternative molecular approach is to focus on known genes that affect the biochemical pathways most likely to be involved in a disease. This type of "candidate gene" study may require a very large sample of patients and healthy subjects for comparison. Julie Williams and I recently coordinated a study involving seven European centres and more than 1,500 research subjects. One of the genes we investigated was called 5HT2a, a receptor for one of the brain's chemical messengers, serotonin. We found that a particular genetic variant conferred a small but significant risk of schizophrenia. This is of particular interest because the newer generation of drugs that relieve schizophrenic symptoms block 5HT2a receptor activity.

³ÉÈËVRÊÓƵ

Critics of genetic research say that it offers only deterministic explanations of behaviour. Saying genes are involved, the argument goes, is the same as saying that social influences play no part or, worse still, that abnormal behaviour is fixed and immutable. This is far from true. Modern medications can treat depression as well as schizophrenia, though there is room for improvement. An obvious way forward is to use molecular genetics to identify the specific brain targets that improved drugs can be designed to hit. Drugs companies are very interested in the genetics of common diseases.

A century ago, the young Sigmund Freud launched the idea that the study of dreams was the royal road to the unconscious and to understanding the springs of human behaviour. I have little doubt that had he been born 100 years later he would be exploring a rather different avenue in his university's genetics lab.

Peter McGuffin is professor of psychological medicine at the University of Wales College of Medicine.