Thursday, April 23, 2009

Who doesn't 'believe' in genes? Who does believe in miracles?

Disputes in science, as in any other field, can become polarizing and accusatory. Skepticism is almost by definition the term applied, usually in a denigrating way, to a minority view. Minority views are often if not usually wrong, but history shows that majority views can be similarly flawed (as we mentioned in our post of April 6). Indeed, major scientific progress occurs specifically when the majority view is shown to be incorrect in important ways.

It is sometimes said, or implied, about those who doubt some of the statements made about mapping complex traits by GWAS (see previous posts) and other methods, that the skeptics 'don't even believe in genes!' The word 'believe' naturally comes to the tongue when characterization of heretics is afoot, and reflects an important aspect of majority views (including the current basic theory in any science): they are belief systems.

In genetics, we know of nobody who doesn't 'believe' in genes. The question is not one of devil-worship by witches like those Macbeth met on a Scottish heath. The question is about what, where, and how genes work and how that is manifest in traits we in the life sciences try to understand. Not to believe in genes would be something akin to not believing in molecules, or heat.

In the case of human genetics, genes associated with and/or responsible for hundreds of traits including countless diseases, have been identified (you can easily read about them in OMIM and elsewhere). Many of these are clearly understandable as the effects, sometimes direct effects, of variation in specific genes.

In some examples, like cystic fibrosis, almost every case of a trait or disease is due to variation in the same gene. In others, such as hereditary deafness, different affected people or families are affected because of the effects of different genes, but it appears that the causal variants are so rare in the population that in each family deafness is only one of them. This is called multiple unilocus causation, and many different genes have been found for such traits (as in the deafness case shown).

Even for more complex traits like, say, diabetes or various forms of cancer, variation at some genes has strong enough effect that standard methods of gene-hunting were able to find them (and, yes, GWAS can find them, too!). BRCA1 and its association with risk of breast cancer is a classic instance of that. But often the results can't be replicated in other studies or populations, or even different families, a problem that has been much discussed in the human genetics literature, including papers written by us.

So, what is at issue these days is not whether genes exist, are important, or are worthy of study (and the same applies to human disease, or studies of yeast, bacteria, insects, flowering plants, or whatever you're interested in). Instead, what is at issue is how genes work in the context of complex traits that involve many interacting genetic and environmental factors.

And in addition to the basic understanding of how genes work in this context (and, incidentally, the rapidly expanding senses of DNA functions besides the usual concept of what a 'gene' is), is the question of how to find the genes and their effects, and what kinds of information that may provide in regard to applications in agriculture or human health.

In the latter case especially, the promise has been that we can predict your health from your DNA. Widely publicized companies are selling this idea in various ways from customer-submitted DNA samples, and some medical geneticists are promising personalized medicine in glowing terms, too.

Indeed, there has long been an effective profession dedicated to this general problem. It's called genetic counseling. Genetic counselors work in a monitored setting, with standardized approaches and ethical procedures in dealing with clients. They systematically collect appropriate clinical and other information. And for known genetic variation associated with disease they, or knowledgeable physicians, can make useful, personalized predictions, explaining options for treatment, family planning, and so on.

It is not a lack of 'belief' in genes, but the opposite--an understanding of genetics--that leads some scientists to question the likely efficacy of this or that proposed direction in health-related research, or in other areas such as criteria for developing evolutionary explanations (i.e., scenarios for past natural selection) for various traits including complex traits like diabetes and even social behavior.

Dispute in science should not be viewed or characterized as if it were the same as dispute in religion....even though both are similar cultural phenomena that often center around accepted theory or dogma. Questions about priorities and dramatic promises for scientific approaches are legitimate and all dogma should be questioned.

Every biologist we know 'believes' in genes. But not all biologists believe in miracles!

No comments: