The Genome's Last Letter

Sixteen years. Three billion dollars. Hundreds of researchers across six countries. We had the source code.

There's just one problem: we still don't know what most of it does.

Chromosome 1 is the largest human chromosome — roughly 247 million base pairs, encoding somewhere around 3,141 genes. Buried within it are instructions for thousands of proteins: genes tied to cancer susceptibility, neurological conditions, metabolic processes, and an almost comically large collection of smell receptors. Humans have around 400 functional olfactory receptor genes, many of them clustered on chromosome 1, evolved over millions of years to detect specific molecules in the environment. Most go unused in a lifetime. Evolution is not efficient; it is experimental.

The completion was genuinely historic. The practical applications arrived fast: personalized medicine, pharmacogenomics, genome-wide association studies linking variants to disease, the ability to read fetal DNA from a blood draw. The genome became a reference map against which billions of individual sequences could be compared. Diagnostics changed. Evolutionary biology changed. Forensics changed.

But then there's the philosophical context, which is stranger than any of the applications.

About 98% of the human genome is noncoding — it doesn't directly encode proteins. For a long time, scientists called it junk DNA. Filler between the real instructions. Evolutionary detritus accumulated over millions of years with no purpose beyond taking up space.

They were wrong, but the error was forgivable. The noncoding regions are harder to read. What they do, it turns out, is something stranger and more interesting: they regulate everything else. Enhancers, promoters, silencers — sequences that control when genes turn on, how strongly, in which tissues, under what conditions. They're not the code. They're the logic that runs the code.

The ENCODE project, launched in 2003, spent a decade cataloguing the biochemical activity of the noncoding genome. Its 2012 findings were controversial but directionally clear: much of the "junk" is doing something. Not all of it — some is evolutionary debris, remnants of ancient viruses and repeated sequences that have shed their function. But the dismissal was wrong. What looks like noise, examined closely enough, turns out to contain structure.

This is the lesson the genome keeps teaching. We read the last letter in 2006. We're still learning the grammar.

i · sources

• Finishing the euchromatic sequence of the human genome — Gregory et al., Nature 441, May 18, 2006