Crystallizing short-read assemblies around lone Sanger reads

New short-read sequencing technologies produce large volumes of 25-30 base paired-end reads. In this paper, we present a sequencing protocol and de novo assembler program (SHORTY) targeted towards such microread data. Our protocol augments short-paired reads using a trivially small number of Sanger reads (only one to three reads per bacterial genome). Still, these “seed reads” enable us to produce significant assemblies using about half the short-read coverage (50-60X) of comparable assemblers, despite our assumption of base error rates at least 10 times that of other groups. SHORTY exploits two new ideas which we believe to be of interest to the shortread assembly community: (1) using single seed reads to crystalize assemblies, and (2) estimating intercontig distances accurately from multiple spanning paired-end reads. Contact: [email protected]