genomic library

Imagine shredding an entire encyclopedia into overlapping snippets, putting each snippet into its own labelled folder, and keeping all the folders together in one cabinet. You can no longer read it cover to cover, but every sentence is in there somewhere, and you can search for the one you need. A genomic library is that cabinet for an organism's DNA.

To build a genomic library, you take the whole genome of an organism, chop it into fragments (with restriction enzymes or by physical shearing), and clone every fragment into a vector — typically a high-capacity vector like a BAC for large genomes. Each clone holds one random piece of the genome, and there are enough clones, with enough overlap, that together they represent every part of the genome at least once. The collection of all these clones, living in bacteria you can freeze and revive, is the library. Crucially, a genomic library contains everything: coding genes, introns, regulatory regions, repetitive DNA, and the vast non-coding stretches between genes — the genome exactly as it is in the chromosome.

A library is a permanent, searchable archive of a genome that you make once and use many times. To find a particular gene you screen the library with a probe — a labelled piece of DNA complementary to your target — which hybridizes to and flags the clone carrying that sequence. Genomic libraries were essential tools before whole-genome sequencing became cheap, and BAC libraries in particular provided the ordered tiling clones that the Human Genome Project sequenced piece by piece.