The authors are Alan Guttmacher MD and Francis Collins MD, PhD, deputy director and director respectively of National Human Genome Research Institute that mapped out a draft sequence of the human genome well ahead of schedule in 2000. Whereas genetics is the study of single genes, genomics, which only entered the lexicon 15 years ago, investigates the functions and interactions of all the genes in the genome, including their causation on different illnesses.

Single-gene conditions such as cystic fibrosis, the authors explain, are the exception rather than the rule. More common are "multifactorial" disorders such as Parkinson's, tuberculosis, HIV, and Alzheimer's, the result of the
interactions of multiple genes and environmental situations. We now know that the human genome comprises 25,000 genes, far fewer than originally thought, and mutations known to cause disease have been identified in about 1,000.

Genes are distributed over 23 pairs of chromosomes, and "express" or turn on or off the proteins that are responsible for everything from digesting food to transmitting nervous signals. This is accomplished through DNA spelled out in seemingly endless four-chemical sequences (A,T,C,G in various arrangements) segmented along three-chemical amino acid bases called "codons."

"A point mutation" involves the equivalent of a genetic typographical error, say a C where an A should be. Since 98.5 percent of genes do not code for proteins, most typos pass unnoticed. But then there are "functional mutations" that can range from causing Huntington's to inducing a bad reaction from eating fava beans. By the same token, mutations can also decrease the risk of disease, including one that makes certain people almost completely resistant to HIV type 1.

The illnesses that we have found genes for are highly "penetrant," ie their mutations lead to diseases in a fairly large number of people who have them. Breast cancer, diabetes, and Parkinson's are examples. Their "prevalence" across the general population, however, is slight, one in several hundred to one in several thousand. On the other side of the coin are genes less highly penetrant but much more prevalent, such as those that increase the risk of colorectal cancer and thrombosis.

Two unrelated people share more than 99.9 percent of their DNA sequences, but since three billion base pairs constitute the human genome, we vary at millions of bases. Researchers are looking to catalogue these variants, called "single-nucleotide polymorphisms" (SNPs) and correlate them with specific "phenotypes," or persons with certain clinical traits. An extension of this effort is to correlate "haplotypes," ie a group of variants that may be inherited together. The authors don't say it, but it is clear that these
are the types of cross-checks that will be used to tease out our thus-far elusive mood genes.

After that, we may be able to find treatments that work rather than half-work. As the authors conclude: "Genomics, which has quickly emerged as the central basic science of biomedical research, is poised to take center stage in clinical medicine as well." Here's hoping that psychiatry is part of that change.