Shallow Shotgun Sequencing as a Strategy for Finding Coding Exons

Abstract

Narrowing candidate genomic segments down to a size compatible with contig assembly by shotgun sequencing is a limiting step in the identification of mammalian genes by positional cloning. A 6 to 8 fold redundancy of sequencing (coverage) is usually required and, most often, does not alleviate the need for directed approaches for closing the remaining gaps. Once a complete sequence is obtained, computer analyzes are used to locate candidate exons, usually spanning less than 10% of the genomic sequence. Here, I propose an alternative strategy in which the need for contig assembly - and the high coverage it imposes - is removed. This strategy takes advantage of the fact that: i) mammalian coding exons are on average much smaller than individual sequencing runs, and ii) computer methods to identify coding regions only depend on local sequence information. In this context of “exon hunting” (opposed to genomic sequencing per se), I show that a 2 to 3 fold sequencing coverage is indeed sufficient to locate most candidate exons within genomic fragments, the size of which is now limited only by the available sequencing power. This strategy can be fully automated as it involves a single experimental technique and real-time computer analyzes of the data.