Physical mapping of complex genomes by cosmid multiplex analysis.

A rapid and powerful approach for linking individual clones of a cosmid library and the assembly of a large physical map is presented, which depends on the simultaneous analysis of many cosmid clones for overlapping regions. This method uses cosmid vectors that contain endogenous bacteriophage T3 and T7 promoters to allow for the identification of overlapping clones through the synthesis of end-specific RNA probes. A genomic library is constructed and organized as an ordered matrix such that each clone is assigned an identifying coordinate. DNA from mixtures of cosmid clones is pooled such that each pool contains only one common member with any other pool, RNA probes are prepared from mixtures of cosmid clones, and groups of clones overlapping with the constituents of the mixtures are determined by hybridization. Pooled probes are most simply prepared by grouping clones according to the rows and columns of the library matrix. The pairwise comparison of data generated by the hybridization of mixed probes can be decoded by using simple algorithms that predict the order and linkage of all clones in the collection and organize them into predicted contigs. To demonstrate the feasibility of multiplexed analysis of cosmids, a genomic library was prepared from a mouse-human somatic cell hybrid that contains a portion of the long arm of human chromosome 11. Preparation, arrangement on a matrix, and analysis of pooled cosmid clones from this collection resulted in the detection of 1099 linked pairs of cosmids, which could be assembled into 315 contigs. Thus, with a minimal amount of effort, a substantial portion of this genomic region has been linked in multiple overlapping contigs. This method may have practical applications in the large-scale mapping and sequencing of mammalian genomes.