Construction of Mammalian Genomic Libraries Using λ Replacement Vectors

The ideal genomic library should consist of a series of clones containing overlapping sequences that are representative of the entire genome. Such an ideal state can be approached by cutting the DNA randomly and cloning large pieces of this DNA into a suitable vector (1). The DNA can be cleaved either by mechanical shearing, in which case the DNA is fragmented in a truely random fashion, but with the introduction of problems associated with blunt end ligation, or better, by partial digestion with a restriction endonuclease such as Mbol or Sau3A. These enzymes recognize four base sequences that are predicted to occur on average (although in practice this is not the case) every 256 bases, and hence digestion results in cleavage of the DNA in a pseudorandom manner.