Sequencing strategies

Introduction
Whole genome shotgun sequencing
Clone-by-clone sequencing

Clone-by-clone sequencing

A different strategy, called “clone-by-clone” or hierarchic sequencing, has been adopted by several international consortia for the sequencing of the human genome and other large genomes (rice, Arabidopsis, roundworm, etc.). The assembly of overlapping reads is no longer performed at the scale of the ensemble of the genome, but at the scale of large genomic fragments (wrongly called “clones”), which are first ordered on a map. This compartmentalization reduces the difficulties due to repetitive sequences, and also makes it possible to target the “finishing” work to precise regions. This finishing step is indispensable in the perspective of an exhaustive and precise annotation of the sequence.

On the other hand, a major mapping effort is needed to put the large genomic fragments (see the Mapping section). This work requires diverse resources: markers from physical and genetic maps, assembly of large overlapping fragments based on their restriction enzyme profile (“fingerprint”), and finally, the end sequences of the large fragments. This last resource may be used as a step-by-step method to construct the map of large fragments; as the sequencing progresses, one can choose overlapping fragments which are least redundant for sequencing.

With a volume of sequences to be read equal to 5 or 6 times the size of a genome, such as that of humans (5X coverage), it is possible to assemble a “draft” clone-by-clone sequence in which several dozen contigs are obtained for each clone. These initial contigs are generally not ordered or oriented. By increasing the coverage to 10X, longer and fewer contigs are obtained, and ordered and oriented thanks to paired reads or other information. This must be followed by fastidious polishing and finishing steps, to ensure less than one error per 10,000 nucleotides, and to fill the residual gaps by working in the specific local regions.