Browsing by Author "Subramanian, Poorani"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Detecting and Correcting Errors in Genome Assemblies(2010) Subramanian, Poorani; Yorke, James A; Applied Mathematics and Scientific Computation; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Genome assemblies have various types of deficiencies or misassemblies. This work is aimed at detecting and correcting a type of misassembly that we call Compression/Expansion or CE misassemblies whereby a section of sequence has been erroneously omitted or inserted in the assembly. Other types of deficiencies include gaps in the genome sequence. We developed a statistic for identifying Compression/Expansion misassemblies called the CE statistic. It is based on examining the placement of mate pairs of reads in the assembly. In addition to this, we developed an algorithm that is aimed at closing gaps and validating and/or correcting CE misassemblies detected by the CE statistic. This algorithm is similar to a shooting algorithm used in solving two-point boundary value problems in partial differential equations. We call this algorithm the Shooting Method. The Shooting Method finds all possible ways to assemble a local region of the genome contained between two target reads. We use a combination of the CE statistic and Shooting Method to detect and correct some CE misassemblies and close gaps in genome assemblies. We tested our techniques both on faux and real data. Applying this technique to 22 bacterial draft assemblies for which the finished genome sequence is known, we were able to identify 5 out of 8 real CE misassemblies. We applied the Shooting Method to a de novo assembly of the Bos taurus genome made from Sanger data. We were able to close 9,863 gaps out of 58,386. This added 8.34 Mbp of sequence to the assembly, and resulted in a 7 % increase of N50 contig size.Item A whole-genome assembly of the domestic cow, Bos taurus(2009-04-29) Zimin, Aleksey V; Delcher, Arthur L; Florea, Liliana; Kelley, David R; Schatz, Michael C; Puiu, Daniela; Hanrahan, Finnian; Pertea, Geo; Van Tassell, Curtis P; Sonstegard, Tad S; Marcais, Guillaume; Roberts, Michael; Subramanian, Poorani; Yorke, James A; Salzberg, Steven LBackground: The genome of the domestic cow, Bos taurus, was sequenced using a mixture of hierarchical and whole-genome shotgun sequencing methods. Results: We have assembled the 35 million sequence reads and applied a variety of assembly improvement techniques, creating an assembly of 2.86 billion base pairs that has multiple improvements over previous assemblies: it is more complete, covering more of the genome; thousands of gaps have been closed; many erroneous inversions, deletions, and translocations have been corrected; and thousands of single-nucleotide errors have been corrected. Our evaluation using independent metrics demonstrates that the resulting assembly is substantially more accurate and complete than alternative versions. Conclusions: By using independent mapping data and conserved synteny between the cow and human genomes, we were able to construct an assembly with excellent large-scale contiguity in which a large majority (approximately 91%) of the genome has been placed onto the 30 B. taurus chromosomes. We constructed a new cow-human synteny map that expands upon previous maps. We also identified for the first time a portion of the B. taurus Y chromosome.