pervasive transcription of the genome
/ per-VAY-siv tran-SKRIP-shun /
When the human genome was first read, a tidy picture was expected: genes are islands of meaning in a sea of silent, junk DNA, and only those genes get transcribed into RNA. Then biologists looked at all the RNA a cell actually makes, and the picture fell apart. Far more of the genome is copied into RNA than anyone predicted — most of it, by some measures. This is pervasive transcription.
Concretely, the protein-coding genes — the part of the genome we long thought of as the only working part — make up only a small slice, around 1 to 2 percent of human DNA. Yet a large fraction of the genome shows up as RNA at some time or place: long non-coding RNAs, microRNA precursors, antisense transcripts read off the opposite strand of a gene, RNAs from old transposons, and many short and obscure transcripts. The cell is not transcribing a few neat genes and ignoring the rest; it is reading huge stretches of its DNA, on both strands, producing a teeming and varied population of RNA called the transcriptome.
What this means is genuinely debated, and honesty matters here. On one hand, it overturned the old assumption that non-coding DNA is dead weight: we now know it harbors regulatory RNAs and elements that matter. On the other hand, some of this transcription is likely biological noise — leaky, low-level copying with no function. The honest summary is the one this field has come to: the genome is pervasively transcribed, RNA is far more than a mere messenger and is a regulator and catalyst in its own right, yet not every transcript is meaningful, and sorting the functional from the incidental is among the central tasks of modern genomics.
Large surveys of the transcriptome find RNA copied from a majority of the human genome at some point, even though protein-coding exons are only about 1 to 2 percent of the DNA — so most of the RNA a cell makes does not encode any protein.
Far more of the genome is transcribed into RNA than codes for protein — most RNA is non-coding.
Pervasive transcription does not mean all that RNA is functional — that is an overstatement the field actively guards against. The honest position is that much non-coding RNA is functional and much remains unproven; abundance of transcription is a starting question, not an answer.