Skip to Main Content

Tailor-made medicines are within our reach, Collins tells genomics conference

Yale Medicine Magazine, 2002 - Autumn


Within two decades a new generation of highly effective designer drugs will spring from our improved understanding of the human genome, according to the director of the National Human Genome Research Institute, Francis S. Collins, M.D., Ph.D. ’74, FW ’84.

As scientists delve more deeply into the human genome, they are developing faster and cheaper techniques to identify individual genetic variations, Collins told the audience attending the fifth annual Pharmacogenetics and Medicine Lectures sponsored by Genaissance Pharmaceuticals in Harkness Auditorium in June. Locating these genetic polymorphisms is crucial to making drugs more effective and preventing harm, because genetic differences lead to different drug responses. Collins believes that by about 2010 information about individual variability will allow doctors to choose medications that best fit the individual’s genetic makeup. “Doctors will have to get used to determining the genotype before writing the prescription,” said Collins, who earned his doctorate in physical chemistry and did a postdoctoral fellowship in genetics at Yale.

Identifying variation is still painstaking, he said, because “the genome is a big place.” Searching for polymorphisms by typing the whole genome for each individual in a study is too slow and far too expensive to be practical, even at the current cost of 50 cents per genotype, said Collins. If an individual’s genome contains about 10 million places where variations are likely (out of 3 billion base pairs), then scanning each genome for the most common variation for a sample of 1,000 people with a disease and 1,000 controls would cost $10 billion.

Researchers are developing shortcuts. One explores the relationship between variants in the genome and their neighbors. Once such a “haplotype map” is developed, this method would reduce the number of variants that have to be tested by a factor of about 40. Another promising shortcut for identifying disease-linked polymorphisms is to pool DNA samples from a large number of individuals, then compare the pooled genomes of people with a disease with the pooled genomes of controls. In Collins’ lab at the NIH, this pooling method has proven accurate to within 3 percent and is far cheaper than analyzing each individual’s genome separately. Collins estimated that a combination of the haplotype map, the pooling method and an anticipated drop in the cost of genotyping would put the price of a study of 1,000 affected people and an equal number of controls at about $50,000.

Given the public’s concern about potential misuse of genetic information, Collins said the nation urgently needs laws barring genetic discrimination and urged audience members to get involved in this debate. He predicted that primary care providers, especially nurses, would play a critical role in educating and counseling patients about how their health care might be affected by advances in human genetics.

A list of links to websites on the human genome can be found at A poster of the genome is available at

Previous Article
Magazine gives students an outlet for creative work and a way to “unpack experiences”
Next Article
For some in public health, changes to chicken have been foul deeds, indeed