The Genome Debate

Singapore has invested heavily in modern biomedical sciences including genomics, so when Prof Steve Jones, a highly respected geneticist at University College London, dissed a major (and expensive) type of genomics study in Britain's Daily Telegraph, it grabbed the attention of my editor. At his request, I supplied the article below.

In the closing paragraph, I allude to a public feud that got some ink a few years ago. The spat was between Mr. Philip Yeo, then the Director of A*STAR, and Dr. Lee Wei Ling who is the Director of the National Neuroscience Institute. The exchange between the two was carried in Singapore's newspapers, and was granted extra titillation quotient by Mr. Yeo's most un-Singaporean bluntness, and Dr. Lee's blue-blood pedigree. (Dr. Lee is the daughter of the nation's founder, Mr. Lee Kwan Yew.)

Dr. Lee decried the vast sums of money being funneled toward molecular genetics studies at the expense of other clinical issues that were just as pressing. The example she gave was head injuries, hence my closing comment.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

I once got a book of poetry as a birthday gift. When I couldn’t find a poem listed in the table of contents, I realized the book was missing about 70 pages, and about the same number of pages were duplicated.

Don’t put the newspaper down. I promise not to talk about poetry. The real subject involves DNA, genes, genomes, and a recent controversy that has pitted some big names in science against each other; people like Professor Steve Jones of University College London and Dr. Mark Walport, Director of the enormous UK-based charity Wellcome Trust. The controversy reverberates in Singapore, given our extensive investment in genetic research and some high-profile figures on opposite sides of the divide, including Dr. Lee Wei Ling, Director of the National Neuroscience Institute, and Dr. Edison Liu, Director of the Genome Institute of Singapore.

The controversy focuses on whether the vast sums of money invested in one avenue of genetic research, broadly known as genomics, have yielded a reasonable bang for the buck.

To understand the debate, we should have a sense of what we mean by the terms genome and genomics. Genes are made of DNA. Because the 4 building blocks of DNA -- abbreviated A, G, C and T – are arranged in a string one after the other, we often use the analogy of the sentence. Every gene is a sentence composed of four letters repeated in a way that is unique for each gene.

If genes are sentences, then genomes are books. The complete set of DNA, carried by each cell of your body, is known as the human genome. It is our book of life.

Each of us has our own copy of this book (the genome), but my copy is not identical to yours. There are small changes and mistakes in each person’s copy of the book. Some mistakes are shared by many books, while others are rare. Some books also have pages repeated or missing, like my poetry book

A major goal of genomics (the study of genomes) is to find which mistakes are responsible for disease. For example, we could ask, “Which mistakes are responsible for increased risk for diabetes?” But to determine this, we would have to compare the genomes of healthy versus diabetic individuals.

This creates a problem. It has been very expensive, time-consuming, and labour-intensive to read even one complete genome. But it takes the comparison of thousands of genomes to locate the mistakes that are the cause of a disease. So how do we get the answer minus the expense?

Well, there’s a trick. Instead of reading the entire genome – the whole book – from thousands of individuals, scientists have instead just sampled every page. This way they have built up a collection of mistakes that might be found on any given page. For example, they may have found 3 mistakes that can occur on page one. Many people will have the first mistake on page one in their copy of the book. Others will have the second. Still others, the third.

Now researchers can get a much quicker and cheaper answer to a different question: “Which kind of page one do diabetics have?” And they can do this for every page in the book. Once they are confident that diabetics have, say, a different kind of page 204, they can be reasonably sure that one of the sentences (genes) on that page is the actual culprit. They can then focus their efforts on getting the full sequence of just that page, and pinpoint the change that causes diabetes. It’s like narrowing the criminal’s whereabouts to one block before starting a door-to-door to search. The technical name for this approach is Genomewide Association Study (GWAS).

Fantastic, right? So what’s the problem? First of all, I have to clarify that “cheaper” doesn’t ‘mean cheap. A recent GWAS funded by Wellcome Trust involved 120,000 patients and cost many millions of dollars. So, if that kind of money – much of it raised through taxes and charities – is disappearing into GWASs, there is an expectation that the research should deliver.

And there’s the rub. Some scientists feel the work has paid off and should be continued. Some feel it has been a good ride, but now a strategic change is required. And still others feel that the whole thing has been a wasteful much ado about nothing.

Prof Jones has caught international attention for a letter to the Daily Telegraph in which he laments the “waste” of money on GWASs. He argues that GWASs have identified very few genes that play a major role in human disease. Instead, many implicated genes make only small contributions, and there are scores or hundreds of such genes for each disease. You can develop a drug to help someone if their disease is caused by one defect. But you cannot do so if it is caused by hundreds.

The quibble with Prof Jones is that he does a good job of dramatizing the disappointments of the GWAS approach without giving sufficient weight to its successes, and without genuinely evaluating the future for genomics research.

First, the successes, as pointed out by Dr. J. N. Hirschhorn in a recent commentary: GWAS has provided major breakthroughs in our understanding of a form of adult blindness (age-related macular degeneration), Crohn’s disease (a debilitating inflammatory bowel disease), the determinants of height (relevant to problems of child development). And just this week we have learned that we can add autism to the list.

Second, even where GWAS does not reveal a gene with a major effect, it can tell us which biological process is at fault. This is beautifully validated by the observation that GWASs have identified genes already known to be effective targets of drugs currently on the market.

Third, as argued by Dr. David Goldstein of Duke University, the next-generation approach in the genomics arsenal is, in fact, whole-genome sequencing of healthy and sick individuals. Yes, I know I said it was too slow and expensive. But that was then, and this is now. It took years to complete the first human genome. Now, one person’s genome can be sequenced in a day. Yes, it’s still expensive, but the time and cost are dropping dramatically. With whole-genome sequencing, the key genes will stand out like an ang moh in Ang Mo Kio.

Last, it is important to note what Prof Jones does not disparage. He does not in any way suggest that molecular genetics, molecular biology, drug discovery and other areas of genomics – the kind of research in which Singapore has invested so handsomely – is a waste of money. His beef is strictly with GWAS and its offspring. Of course, if you ask me whether Singapore has done the right thing – I’m glad you asked – my answer is an unqualified yes.

This is not to devalue other areas of research in human health. Whether my little daughter succumbs to a genetic disease or suffers a head injury, I want well-funded research to bring her the best treatment possible. After all, I have a lot of bedtime stories to read her, preferably from books with all their pages.


Mark Featherstone is a Professor at the School of Biological Sciences, Nanyang Technological University. He acknowledges a useful article, “Genes show limited value in predicting diseases”, by Nicholas Wade of the New York Times (16 April, 2009).